NAVAL

POSTGRADUATE
SCHOOL
MONTEREY, CALIFORNIA

THESIS

FACING REALITY: THE BENEFITS

AND CHALLENGES OF FACIAL RECOGNITION
FOR THE NYPD

by

Anthony M. Carter

September 2018

Co-Advisors: Lauren Wollman (contractor)

Carolyn C. Halladay
Approved for public release. Distribution is unlimited.
THIS PAGE INTENTIONALLY LEFT BLANK
 Form Approved OMB
REPORT DOCUMENTATION PAGE No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing
instruction, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of
information. Send comments regarding this burden estimate or any other aspect of this collection of information, including
suggestions for reducing this burden, to Washington headquarters Services, Directorate for Information Operations and Reports, 1215
Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction
Project (0704-0188) Washington, DC 20503.
1. AGENCY USE ONLY 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED
(Leave blank) September 2018 Master's thesis
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS
FACING REALITY: THE BENEFITS AND CHALLENGES OF FACIAL
RECOGNITION FOR THE NYPD
6. AUTHOR(S) Anthony M. Carter

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING

Naval Postgraduate School ORGANIZATION REPORT
Monterey, CA 93943-5000 NUMBER
9. SPONSORING / MONITORING AGENCY NAME(S) AND 10. SPONSORING /
ADDRESS(ES) MONITORING AGENCY
N/A REPORT NUMBER

11. SUPPLEMENTARY NOTES The views expressed in this thesis are those of the author and do not reflect the
official policy or position of the Department of Defense or the U.S. Government.
12a. DISTRIBUTION / AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE
Approved for public release. Distribution is unlimited. A
13. ABSTRACT (maximum 200 words)
Facial recognition technology (FRT) and license plate readers (LPRs) are comparable technologies that
share similar benefits and challenges. Many of the challenges associated with LPRs have already been
mitigated, resulting in citizen acceptance and expanded use. Recent advancements in FRT, such as its use in
real time, create new opportunities to leverage the technology for increased public safety. To what extent are
LPRs and FRT analogous, and how can the use of LPRs by the New York City Police Department (NYPD)
provide a roadmap for public support of real-time FRT? This thesis examines benefits and challenges that
may arise if the NYPD considers using real-time FRT in the New York City subway system. Through
comparative analysis, this thesis determines that real-time FRT could help law enforcement deter terrorism,
prevent violent crime, identify wanted individuals, find missing persons as well as assist in mental health
situations and post-event investigations. Real-time FRT can help the NYPD meet its mission by reducing
fear, increasing resiliency, and adding a layer of protection for citizens riding in the New York City subway
system.

14. SUBJECT TERMS 15. NUMBER OF

facial recognition, facial recognition technology, FRT, real-time facial recognition PAGES
technology, New York City Police Department, NYPD, license plate readers, police 123
technology, LPR 16. PRICE CODE
17. SECURITY 18. SECURITY 19. SECURITY 20. LIMITATION OF
CLASSIFICATION OF CLASSIFICATION OF THIS CLASSIFICATION OF ABSTRACT
REPORT PAGE ABSTRACT
Unclassified Unclassified Unclassified UU

NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89)

Prescribed by ANSI Std. 239-18

i
THIS PAGE INTENTIONALLY LEFT BLANK

ii
 Approved for public release. Distribution is unlimited.

FACING REALITY: THE BENEFITS AND CHALLENGES OF FACIAL

RECOGNITION FOR THE NYPD

Anthony M. Carter
Deputy Inspector, New York City Police Department
BS, The State University of New York, Empire State College

Submitted in partial fulfillment of the

requirements for the degree of

MASTER OF ARTS IN SECURITY STUDIES

(HOMELAND SECURITY AND DEFENSE)

from the

NAVAL POSTGRADUATE SCHOOL

September 2018

Approved by: Lauren Wollman

Co-Advisor

Carolyn C. Halladay
Co-Advisor

Erik J. Dahl
Associate Chair for Instruction,
Department of National Security Affairs

iii
THIS PAGE INTENTIONALLY LEFT BLANK

iv
 ABSTRACT

Facial recognition technology (FRT) and license plate readers (LPRs) are
comparable technologies that share similar benefits and challenges. Many of the
challenges associated with LPRs have already been mitigated, resulting in citizen
acceptance and expanded use. Recent advancements in FRT, such as its use in real time,
create new opportunities to leverage the technology for increased public safety. To what
extent are LPRs and FRT analogous, and how can the use of LPRs by the New York City
Police Department (NYPD) provide a roadmap for public support of real-time FRT? This
thesis examines benefits and challenges that may arise if the NYPD considers using
real-time FRT in the New York City subway system. Through comparative analysis, this
thesis determines that real-time FRT could help law enforcement deter terrorism, prevent
violent crime, identify wanted individuals, find missing persons as well as assist in
mental health situations and post-event investigations. Real-time FRT can help the NYPD
meet its mission by reducing fear, increasing resiliency, and adding a layer of protection
for citizens riding in the New York City subway system.

v
THIS PAGE INTENTIONALLY LEFT BLANK

vi
 TABLE OF CONTENTS

I. INTRODUCTION..................................................................................................1
A. RESEARCH QUESTION .........................................................................2
B. OVERVIEW OF FACIAL RECOGNITION TECHNOLOGY............2
C. UNINTENDED CONSEQUENCES OF BODY-WORN
CAMERAS .................................................................................................5
D. LITERATURE REVIEW .......................................................................11
1. Privacy versus Security ...............................................................11
2. Transparency................................................................................19
E. RESEARCH DESIGN .............................................................................23
F. CHAPTER OUTLINE.............................................................................24

II. LICENSE PLATE READERS ...........................................................................25

A. OVERVIEW .............................................................................................25
1. History of License Plate Reader Technology.............................25
2. How License Plate Readers Work ..............................................28
B. BENEFITS OF LPR TECHNOLOGY ..................................................30
1. Crime Prevention Benefits ..........................................................30
2. Law Enforcement and Public Benefits.......................................34
C. CHALLENGES ........................................................................................37
1. Privacy ..........................................................................................38
2. Accuracy .......................................................................................41
3. Transparency and Public Trust..................................................43
D. CONCLUSION ........................................................................................46

III. FACIAL RECOGNITION TECHNOLOGY ....................................................49

A. OVERVIEW .............................................................................................49
B. BENEFITS OF FRT ................................................................................50
1. Crime Prevention Benefits ..........................................................53
2. Law Enforcement and Public Benefits.......................................59
C. CHALLENGES ........................................................................................64
1. Privacy ..........................................................................................65
2. Accuracy .......................................................................................73
3. Transparency and Public Trust..................................................75
D. CONCLUSION ........................................................................................77

IV. CONCLUSION ....................................................................................................79

vii
LIST OF REFERENCES ................................................................................................85

INITIAL DISTRIBUTION LIST .................................................................................101

viii
 LIST OF FIGURES

Figure 1. Notional Large-Scale, Integrated Departmental LPR Setup ......................29

Figure 2. Algorithm for Matching Facial Images......................................................51

Figure 3. How FRT Systems Generally Work ..........................................................53

Figure 4. The Gateless Gate Line Concept Using FRT .............................................62

ix
THIS PAGE INTENTIONALLY LEFT BLANK

x
 LIST OF TABLES

Table 1. Comparison of Benefits of LPRs, FRT, and BWCs ..................................82

Table 2. Comparison of Challenges with LPRs, FRT, and BWCs ..........................83

xi
THIS PAGE INTENTIONALLY LEFT BLANK

xii
 LIST OF ACRONYMS AND ABBREVIATIONS

ACLU American Civil Liberties Union

BWC body-worn camera
CCTV closed-circuit television
CPOP community patrol officer program
CUFPR Communities United for Police Reform
DAS domain awareness system
DHS Department of Homeland Security
DMV Department of Motor Vehicles
FBI Federal Bureau of Investigation
FIS Facial Identification Section
FRT facial recognition technology
ICE Immigration and Customs Enforcement
INTERPOL International Police
KRVS known-risk vehicle stop
LMSI Lower Manhattan Security Initiative
LPR license plate reader
MTA Metropolitan Transit Authority
NYPD New York City Police Department
OPM Office of Personnel Management
SQF stop, question, and frisk
VRA vehicle-ramming attack

xiii
THIS PAGE INTENTIONALLY LEFT BLANK

xiv
 EXECUTIVE SUMMARY

Facial recognition technology (FRT) is a biometric technology that—if the New

York City Police Department (NYPD) implements it in the New York City subway
system—could have significant benefits of preventing violent crime, deterring terrorism,
helping investigate past crimes, locating missing persons, providing assistance to
individuals with special needs, and integrating with other technology platforms to allow
for greater efficiencies in policing. Since the NYPD successfully uses license plate reader
(LPR) technology to increase public safety with an existing legal and policy framework
governing its use, and because FRT is similar to LPR technology—in terms of law
enforcement use of the technology as well as benefits and challenges—this thesis presents
a comparative analysis of the two technologies.

From this analysis, the following questions can be answered: To what extent are
LPRs and FRT analogous, and how can the NYPD’s use of LPRs provide a roadmap for
public support of real-time FRT? Currently, the NYPD does not use automatic or real-time
FRT, but specially trained detectives assigned to the Facial Identification Section use FRT
to investigate past crimes and assist detectives in the field. Other law enforcement agencies
outside the United States, however, are using or evaluating the effectiveness of real-time
FRT, which has significantly improved in recent years, in part because of the development
of high-definition video, advancements in storage capabilities, and the ability to evaluate
faces in real time. Modern FRT systems can also recognize an individual with varying
facial expressions. Therefore, a person can be accurately identified in a facial recognition
database, even if the facial expression is different from the original image contained in the
database.

Communities benefit from LPR technology, despite privacy concerns and the
contention of critics, such as the American Civil Liberties Union, which assert that law
enforcement agencies should not be able to collect and store information on law-abiding

xv
citizens who are not suspected of criminal wrongdoing. 1 This thesis proposes that although
privacy considerations may exist, the benefits of LPR technology outweigh these concerns.
LPRs and FRT are similar types of technologies that perform similar functions. They both
scan the image of an unknown variable and attempt to match it against a known variable
contained in a database, collecting all information—without bias—on license plates or
persons. Furthermore, both technologies use “hot lists,” which compare the scanned image
against information or images contained in local, state, and federal databases. Computer
software then compares the scanned image against collected data or images. Finally, the
information, such as images of a stolen vehicle or wanted person, received from both
technologies requires human verification. Like LPR technology, FRT—especially when
used in real time—has significant benefits and can act as a force multiplier for limited
NYPD resources in crowded environments such as the New York City subway system.

The NYPD has developed sound policies with LPRs and has already mitigated the
challenges to minimize potential harm due to misuse and violation of civil liberties. In
addition, LPRs are generally acknowledged as a common and publicly accepted law
enforcement technology, largely in part because the NYPD proactively addressed many of
the risks. The NYPD addressed LPR data collection, retention, and sharing through robust
and clear policies. In addition, the police department addressed potential misuse of LPRs
to eliminate their ambiguities and clearly define acceptable practices. Through comparative
analysis, this thesis determines that real-time FRT could help law enforcement deter
terrorism, prevent violent crime, identify wanted individuals, find missing persons, and
assist in mental health situations and post-event investigations. This thesis also addresses
the litany of challenges in the use of FRT—privacy concerns as well as false positives,
false negatives, intentional circumvention of real-time FRT, and law enforcement
misappropriations—and identifies the concerns over how law enforcement collects, shares,
and disseminates personal information obtained from facial recognition software. This
thesis concludes that real-time FRT can help the NYPD meet its mission by reducing fear,

1 Ben Eisler, “ACLU Concerned Automatic License Plate Readers May Invade Privacy,” WJLA
News, July 30, 2012, http://wjla.com/news/local/aclu-concerned-red-light-cameras-may-invade-privacy-
78301.
xvi
increasing resiliency, and adding a layer of protection for citizens riding in the New York
City subway system.

xvii
THIS PAGE INTENTIONALLY LEFT BLANK

xviii
 ACKNOWLEDGMENTS

I owe my deepest debt and gratitude to my wife and amazing children. I can never
thank all of you enough for your love, support, and understanding over the last 18 months.
Also, to Mom and Dad, thank you for your guidance, continuous support, and love over
the last 40-plus years.

Thank you, Police Commissioner James O’Neill and First Deputy Commissioner
Benjamin Tucker for allowing me to participate in this most prestigious program. Thank
you, Chief Terence Monahan, Deputy Chief Frank Vega, and Assistant Chief Michael
Lipetri, for your continued support during my extended absences throughout this program.
To my advisors, Carolyn Halladay and Lauren Wollman, you both were instrumental
during the thesis process and in furthering the depth of my argument through countless
emails, calls, and text messages. You helped me focus on the importance of this project.
Also, thank you to my writing coaches—Carla Orvis Hunt, Cheryldee Huddleston, and
Noel Yucuis—for all your support and guidance.

I am extremely grateful to the Naval Postgraduate School and the Center for
Homeland Defense and Security for accepting me into this program. I will treasure the
knowledge, the memories, and the excellence that has become a part of who I am.

Thank you to all of my classmates. I will miss being surrounded by some of the
most intelligent and accomplished individuals I have ever met. My greatest experiences in
learning about homeland security came from you. Thank you all for your contributions to
the security and safety of this great country.

Finally, to the men and woman of the New York City Police Department, I thank
you all for what you do each and every day. I am truly grateful to be considered a member
of the greatest police department in the world, and it is an honor to walk among you. Stay
safe, and God bless.

xix
THIS PAGE INTENTIONALLY LEFT BLANK

xx
 I. INTRODUCTION

Law enforcement organizations around the world are using real-time facial
recognition technology (FRT) to identify known terrorists, apprehend wanted perpetrators,
and locate missing persons. Law enforcement agencies in Europe are also piloting real-
time FRT for use in parades, sporting events, and subway systems. For example, the
Metropolitan Police in London recently used real-time FRT at the Notting Hill Carnival to
identify and arrest wanted criminals, quelling the violence that has marred the event in
previous years. 1 Biometric technology is advancing rapidly, allowing law enforcement
agencies to use real-time FRT with any device equipped with a camera, such as cell phones,
fixed cameras, mobile cameras, body-worn cameras (BWCs), and drones. In China, the use
of real-time FRT by law enforcement has proven successful. For example, in Zhengzhou,
facial recognition identified a drug smuggler, and in Wuhu, facial recognition cameras
identified a murder suspect buying food from a street vendor. 2 Such examples from various
parts of the world illustrate that FRT, especially when used in real time, has significant
public benefits, despite several concerns associated with its use.

FRT is a biometric technology that—if considered by the New York City Police
Department (NYPD) for use in the New York City subway system—could have the
significant benefits of preventing violent crime, deterring terrorism, investigating past
crimes, locating missing persons, providing assistance to individuals with special needs,
and integrating with other technology platforms to allow for greater efficiencies in policing.
Currently, a knowledge gap exists in the NYPD and other law enforcement organizations
for successful implementation of real-time FRT. The NYPD needs to examine the benefits,
challenges, cost, and overall effectiveness thoroughly, prior to adopting FRT for real-time
use. The possibility exists that FRT is similar to LPR technology in terms of how law

1 Vikram Dodd, “Met Police to Use Facial Recognition Software at Notting Hill Carnival,” Guardian,
August 5, 2017, http://www.theguardian.com/uk-news/2017/aug/05/met-police-facial-recognition-software-
notting-hill-carnival.
2 Paul Mozur, “Inside China’s Dystopian Dreams: A.I., Shame and Lots of Cameras,” New York
Times, July 9, 2018, https://www.nytimes.com/2018/07/08/business/china-edsurveillance-technology.html.
1
enforcement uses the technology, as well as respective benefits and challenges of both
technologies. The NYPD successfully uses license plate reader (LPR) technology and has
an existing framework that governs its use. In addition, both technologies have similar
crime prevention benefits and are operationally analogous because they can scan an
unknown image and compare it against known faces or license plates contained in a
database.

The NYPD continuously evaluates new technologies that may further its ability to
keep citizens safe, deter crime, and prevent acts of terrorism. Using real-time FRT may
provide additional layers of protection against known criminals and/or terrorist threats.
Controlled access, lighting, environmental factors, and an existing network of cameras
make the New York City subway system an ideal location for the use of real-time FRT.
This thesis examines LPRs and FRT to test the idea that both technologies are in fact
similar, so a similar framework for real-time FRT can be adopted.

A. RESEARCH QUESTION

To what extent are LPRs and FRT analogous, and how can the NYPD’s use of
LPRs provide a roadmap for public support of real-time FRT?

B. OVERVIEW OF FACIAL RECOGNITION TECHNOLOGY

Kelly Gates defines real time as “making the outcome of mediated processes
happen immediately.” 3 Gates adds that FRT, when used in real time, “involves a complex
technical process of detecting faces in video feeds, grabbing them from the image,
segmenting them from the background clutter, applying an algorithm to translate those
images into a digital template or ‘faceprint,’ and then searching that template against
databases of archived photographs.” 4 Real-time identification happens instantly, making
real-time use of FRT an attractive feature for law enforcement. Currently, the NYPD has
specially trained detectives assigned to its Facial Identification Section (FIS), which uses

3 Kelly Gates, “Identifying the 9/11 ‘Faces of Terror,’” Cultural Studies 20, no. 4/5 (September 7,
2006): 426, https://doi.org/10.1080/09502380600708820.
4 Gates, 426.

2
FRT to investigate past crimes and provides resources to detectives in the field. The NYPD
does not use automatic, or real-time, FRT in any capacity. Other law enforcement agencies
outside the United States, however, are using real-time FRT for subway passenger
identification, transportation payment, crime fighting, access control into transit facilities,
and terrorism mitigation. 5

FRT has significantly improved in recent years in part because of the development
of high-definition video and advancements in storage capabilities, pushing vast quantities
of information into the cloud. 6 Modern FRT systems can also recognize an individual with
varying facial expressions—even expressions of anger, disgust, fear, happiness, sadness,
or surprise. 7 An individual can be correctly identified in a facial recognition database
despite having facial expressions different from the original image contained in the
database. In addition, software companies integrate mobile applications to synchronize
with FRT software. 8 Such integrations create new purposes for FRT.

The RAND Corporation defines biometrics as “any automatically measurable,

robust and distinctive characteristic or personal trait that can be used to identify an
individual or verify the claimed identity of an individual.” 9 FRT processes and matches
unique characteristics for identification or authorization. The technology uses a digital or
video camera to detect images of an unknown individual, analyzes the features within the

5 The Shanghai Metro has 367 stations and moves approximately 10.6 million passengers daily.
Shanghai Shentong Metro Group, in conjunction with Ant Financial Services Group, announced that it will
soon install facial recognition at station entrances. Bien Perez, “Shanghai Subway to Use Alibaba Voice
and Facial Recognition Technologies,” South China Morning Post, December 5, 2017,
http://www.scmp.com/tech/enterprises/article/2123014/shanghai-subway-use-alibaba-voice-and-facial-
recognition-systems-ai.
6 Peter Mell and Timothy Grance, The NIST Definition of Cloud Computing: Recommendations of the
National Institute of Standards and Technology, Special Publication 800–145 (Gaithersburg, MD: NIST,
September 2011), 2, http://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-145.pdf.
7 Shubhada Deshmukh, Manasi Patwardhan, and Anjali Mahajan, “Survey on Real-Time Facial
Expression Recognition Techniques,” IET Biometrics 5, no. 3 (September 2016): 162.
8 “Biometrica Announces Release of New SSIN, with Mobile App That Allows Near Real-Time
Facial Recognition,” Biometrica Systems, November 23, 2016, https://biometrica.com/biometrica-
announces-release-of-new-ssin-with-mobile-app-that-allows-near-real-time-facial-recognition/.
9 John D. Woodward Jr. et al., Biometrics: A Look at Facial Recognition (Santa Monica, CA: RAND
Corporation, 2003), 1, http://www.dtic.mil/dtic/tr/fulltext/u2/a414520.pdf.
3
image, and compares them with images of known individuals contained in the database. 10
Two different methods enable the use of FRT. The first method relies on “criteria like the
distance between your eyes, the measurements of your nose, lips and other facial features
and matches them against an existing database.” 11 The second method “looks at points of
interest on the face and tracks how the pixels in a photograph cluster to form a person’s
nose.” 12 Facial recognition is similar to other biometrics, such as fingerprints, iris scans,
and voice recognition, because these technologies begin with an unknown variable, which
they attempt to identify.

Fifty years since the inception of their use, FRT and other biometrics have
integrated with existing technologies, such as cellular phones and digital photo albums, so
people are more accustomed to the technology—which may help the NYPD garner more
public acceptance of FRT’s benefits and efficiencies. Just as technology companies, such
as Google, Amazon, Facebook, and Apple, are leveraging FRT, so, too, are law
enforcement organizations. Law enforcement has begun to apply FRT in ways that contrast
starkly with the technology’s original use of preventing terrorism.

Citizens continue to fear threats of terrorism when moving around public places
and often look to uniformed police officers to protect them from harm. Kelly Gates argues
that automated FRT was born in the wake of 9/11, with the entire country looking for a
“high-tech” method to identify and deter terrorists. 13 Gates also argues that supporters of
biometric technologies used the war on terror to identify new initiatives for redefining
citizen expectations of homeland security. As stated by British sociologist Andrew Barry,
the United States was obsessed with finding any and all available technological solutions. 14

10 “Facial Recognition,” Find Biometrics, Global Identity Management, accessed August 7, 2018,
https://findbiometrics.com/solutions/facial-recognition/.
11 Stacy Higginbotham, “Facial Recognition Freak Out: What the Technology Can and Can’t Do,”
Fortune, June 23, 2015, http://fortune.com/2015/06/23/facial-recognition-freak-out/.
12 Higginbotham.
13 Kelly Gates, “Identifying the 9/11 ‘Faces of Terror,’” 418.
14 Gates, 423.

4
 New York is the largest city in America and is home to one of the most extensive
and safest subway systems in the world. 15 Terrorists, however, have successfully targeted
New York City three times since September 2016, and those wishing to plan attacks
continue to view it as a primary target. In addition to the NYPD’s daily responsibility to
reduce crime, the continuous undertaking to prevent a large-scale terror attack in the
subway system has never been more significant. New York City subway facilities are often
overcrowded, particularly during peak times, making it virtually impossible to screen every
passenger entering the station. 16 Approximately six million people navigate their way
through 472 stations inside thousands of packed subway cars each day. 17 The massive
volumes of people who traverse the New York City subway system daily make trains and
platform areas attractive targets for terrorists.

On the other hand, if the NYPD uses FRT as a real-time solution in the New York
City subway system, several obstacles exist, such as privacy concerns, public trust
challenges, and the overall reliability of the technology. Real-time FRT can become part
of an all-encompassing policing strategy, but its implementation in the New York City
subway system is not the magic bullet for reducing violent crime or preventing a terror
attack. It could be, however, an additional tool for the NYPD to use for the ever-changing
challenges of policing the nation’s largest subway system.

C. UNINTENDED CONSEQUENCES OF BODY-WORN CAMERAS

In major cities around the world, police departments are harnessing BWC
technology to video record police interactions with citizens. A BWC is typically worn by
a police officer for recording the audio and video of encounters, as well as evidence
searches. BWCs allow others to evaluate the actions of a police officer or citizen based on

15 Lisa Anderson, “Exclusive-Poll: New York City Transport Seen as Safest in World for Women,”
Reuters, October 28, 2014, https://uk.reuters.com/article/women-poll-newyork/exclusive-poll-new-york-
city-transport-seen-as-safest-in-world-for-women-idUKL6N0SB4WI20141029.
16 Emma G. Fitzsimmons, “Every New York City Subway Line Is Getting Worse. Here’s Why,” New
York Times, June 28, 2017, https://www.nytimes.com/interactive/2017/06/28/nyregion/subway-delays-
overcrowding.html.
17 “Introduction to Subway Ridership,” Metropolitan Transportation Authority, accessed September 4,
2017, http://web.mta.info/nyct/facts/ridership/.
5
video footage collected from one or more officers. The NYPD is in the process of
completing the implementation of BWCs and will soon equip every police officer,
detective, sergeant, and lieutenant working on the street. 18 Since the implementation of
BWCs in 2014, differing opinions exist on their success and the over-arching benefit to
both citizens and police officers. Advocacy groups are still collectively debating LPR and
BWC use by law enforcement, but BWCs and LPRs serve different purposes and have
separate challenges for law enforcement, and therefore, decision-makers should not draw
comparisons.

BWCs were seen by many as a panacea to hold officers accountable for police use-
of-force encounters and to provide transparency in policing, thereby enabling stronger
community relations between law enforcement and citizens. Their use, however, has
created issues due to the high costs of BWC programs and high-profile controversies
between law enforcement and citizens, due to policy inconsistencies from one police
department to the next. For example, some departments readily release BWC videos to the
public while others are more restrictive in releasing them. BWCs were supposed to remove
the ambiguity of policing, create an environment where policing is transparent, improve
relations between police and the community, and decrease civilian complaints against
police officers and use-of-force incidents.

Various surveillance technologies, such as video cameras, have been around for
over 50 years. In the last two decades, video cameras have become ubiquitous—because
of their price and quality—and often integrated with other devices such as smartphones.
Just as the use of LPR technology by law enforcement has rapidly expanded within a short
period, both law enforcement decision-makers and critics quickly accepted BWC
technology as a panacea to attain greater transparency and accountability between police
officers and citizens. Police-involved use of force—beginning with the death of Eric
Garner in Staten Island, New York, and followed by the heavily scrutinized deaths of
Michael Brown in Ferguson, Missouri, and Freddie Gray in Baltimore, Maryland—set off

18 Rebecca Savransky, “De Blasio: NYPD Planning to Have Body Camera on Every Cop by Year’s
End,” The Hill, January 31, 2018, http://thehill.com/homenews/state-watch/371582-de-blasio-says-nyc-
planning-to-have-every-cop-equipped-with-body-camera.
6
violent protests throughout the country. A citizen observing the confrontation, arrest, and
subsequent death of Eric Garner captured the incident on video. There are often questions
about police-related interactions captured on video and posted on social media sites for
citizens to critique. Sometimes, video footage clearly captures officer misconduct, such as
the police-involved shooting of Walter Scott, an unarmed man following a traffic stop, by
a South Carolina police officer. Other videos capturing police action raise more questions
than answers. In Louisville, Kentucky, BWCs captured the aftermath of a vehicle collision,
resulting in the deaths of Isaiah Basham and Lexi Grey. Questions about whether the
officers were engaged in a vehicle pursuit are inconclusive based on BWC footage. Scott
Greenwood of the American Civil Liberties Union (ACLU) argues that many police
encounters are already recorded in a variety of ways. At a 2013 conference, he spoke in
favor of BWCs:

The average interaction between an officer and a citizen in an urban area is

already recorded in multiple ways. The citizen may record it on his phone.
If there is some conflict happening, one or more witnesses may record it.
Often there are fixed security cameras nearby that capture the interaction.
So the thing that makes the most sense—if you really want accountability
both for your officers and for the people they interact with—is to also have
video from the officer’s perspective. 19

A BWC varies in size, specifications, quality, and durability, just as in

commercially sold video cameras. The location on a police officer’s body where a BWC is
mounted depends on the size and shape of the equipment. The most significant difference
between closed-circuit television (CCTV) cameras and BWCs is the direct involvement of
the police officer and his actions. Passive CCTV recordings are not reviewed or retained
for a prolonged period unless a specific incident necessitates their review and preservation.
BWCs, however, record active engagement between an officer and citizen, capturing up-
close, high-quality audio and video of specific interactions.

There are privacy and community concerns associated with the use of BWCs.
Policy issues for BWCs often involve the effects of privacy, community relations, overall

19 Lindsay Miller and Jessica Toliver, Implementing a Body-Worn Camera Program:

Recommendations and Lessons Learned (Washington, DC: Community Oriented Policing Services, 2014),
1, https://www.justice.gov/iso/opa/resources/472014912134715246869.pdf.
7
cost, court testimony, and concerns raised internally by officers wearing the technology.
Demonstrating the most significant difference between fixed cameras and BWCs, officers
wearing the latter often enter private residences and other locations not accessible to the
public. When officers activate their cameras, their decisions need to conform to the
requirements set forth by their respective departments, while striking a balance between
transparency and individual privacy. 20 Additional privacy concerns with BWCs involve
how the footage is stored, used, and shared—particularly when footage captures images of
a private or personal nature. 21 Sir Hugh Orde, president of the Association of Chief Police
Officers (UK), explains the importance of effective policies involving BWCs:

Legitimacy in policing is built on trust. And the notion of video-recording

every interaction in a very tense situation would simply not be a practical
operational way of delivering policing. In fact, it would exacerbate all sorts
of problems. In the United Kingdom, we’re also subject to human rights
legislation, laws on right to privacy, right to family life, and I’m sure you
have similar statutes. It’s far more complicated than a blanket policy of
“every interaction is filmed.” I think that’s far too simplistic. We have to
give our officers some discretion. We cannot have a policy that limits
discretion of officers to a point where using these devices has a negative
effect on community–police relations. 22

In some states, the law mandates that officers obtain consent from an individual
prior to activating a BWC. For the NYPD, the patrol guide defines specific rules and
regulations that every officer must follow and dictates precisely when an officer should
activate a BWC. Officers must activate their BWCs when ShotSpotter—a technology that
uses sensors and software—identifies gunfire, during interior vertical patrols of New York
City Housing Authority buildings, or in anticipation of a citizen interaction to capture the
incident in its entirety. The NYPD acknowledges, however, that BWCs do not capture all
encounters and may not clearly depict the complete incident. The NYPD prohibits BWC
recordings of interviews with confidential informants or victims of sex crimes, during strip

20 Miller and Toliver, 11.

21 Miller and Toliver, 11.
22 Miller and Toliver, 14.

8
searches, while inside courthouses (except when lodging a prisoner), and inside medical
facilities. 23

Such policies, although required by law, fail to capture some police encounters with
citizens. On June 30, 2017, Henry Bello, a disgruntled doctor, walked into Bronx Lebanon
Hospital armed with an AR-15 assault rifle and ascended to the 16th and 17th floors, where
he began shooting randomly, killing one and wounding six before taking his own life.
Notably, because BWCs do not capture all incidents, no single technology exists that can
address all police and citizen concerns. Privacy laws prevent police officers from recording
inside hospitals and, therefore, BWCs would not have captured the shooting or its
aftermath. Therefore, an assessment of police tactics during this active-shooter incident
could not have been evaluated using BWC video footage.

Psychologists often contend that individuals behave differently when others are
observing their actions. Researchers Munger and Harris examined the behavior of
individuals and found that when people know they are being watched, they are more likely
to exhibit normal and socially acceptable behavior. 24 A 2012 study involving BWCs by
Munger and Harris concluded that officers wearing BWCs had a 60 percent decrease in
use-of-force incidents and an 88 percent decrease in the number of civilian complaints
lodged against them. 25 The strategy behind BWCs is to reduce civilian complaints against
police and officer use-of-force incidents against civilians, increase officer safety, and build
stronger court cases. 26

The overall benefits obtained from the use of BWCs have been advantageous to
both law enforcement and citizens alike. BWCs have reduced civilian complaints against

23 New York City Police Department, Patrol Guide: Command Operations, 212–123 (New York:
NYPD, January 8, 2018), https://www1.nyc.gov/assets/nypd/downloads/pdf/public_information/
body-worn-cameras-patrol-guide.pdf.
24 Kristen Munger and Shelby J. Harris, “Effects of an Observer on Handwashing in a Public
Restroom,” Perceptual and Motor Skills 69 (1989): 733–734, https://kmunger.files.wordpress.com/2007/
09/munger-harris-1989-effects-of-an-observer-on-handwashing-in-a-public-restroom.pdf.
25 Miller and Toliver, Implementing a Body-Worn Camera Program, 14.
26 Paul Drover and Barak Ariel, “Leading an Experiment in Police Body-Worn Video Cameras,”
International Criminal Justice Review 25, no. 1 (March 1, 2015): 80–81, https://doi.org/10.1177/
1057567715574374.
9
police officers, with sharp decreases in police use-of-force incidents and fewer falsified
complaints made by citizens. 27 Internally, law enforcement organizations can use BWC
footage to train and identify problematic tactics or officer behavior. BWC technology is
also vital in reconstructing officer-involved shootings as well as documenting evidence,
court testimony, and consents to search.

In addition to privacy concerns, questions remain as to when officers should be

required to activate their BWCs. Policies vary by department, with some choosing to record
every encounter and others allowing officer discretion of what and when to record. There
are debatable positions by advocates and law enforcement alike on these issues. Officers
from the Sunnyvale Police Department, who participated in a survey, show the negative
impact BWCs have had on officer decision making. The survey found that “over half
(14/23) strongly agreed or agreed with the statement, ‘I feel like I have to follow the letter
of the law when wearing my body-worn camera.’” 28 Furthermore, 15 officers indicated on
the survey that they “felt uncomfortable ‘cutting breaks’ on the street.” 29 Officers felt that
discretion was more difficult when dealing with minor offenses. Officer interviews
“suggested that a reason for this change is that cameras now made it possible for others,
particularly superiors, to examine an officer’s decision retrospectively and assess its
appropriateness given the particular circumstances.” 30 As one officer explained, “The
possibility of increased scrutiny made him more likely to write a traffic ticket, instead of
giving written warnings or just letting someone off with a ‘pep-talk.’” 31 “More than half
of the officers (13/24) reported that wearing a camera influenced how they communicated
with people.” 32 In fact, the officers acknowledged the potential for their words to extend
beyond the individuals involved in an encounter: “Hence, they were more prone to exercise

27 Miller and Toliver, Implementing a Body-Worn Camera Program, 5–7.

28 Marthinus C. Koen, James J. Willis, and Stephen D. Mastrofski, “The Effects of Body-Worn
Cameras on Police Organisation and Practice: A Theory-Based Analysis,” Policing and Society (April
2018): 8, https://www.tandfonline.com/doi/pdf/10.1080/10439463.2018.1467907?needAccess=true.
29 Koen, Willis, and Mastrofski, 8.
30 Koen, Willis, and Mastrofski, 8.
31 Koen, Willis, and Mastrofski, 8.
32 Koen, Willis, and Mastrofski, 8.

10
‘verbal caution,’ which meant avoiding profanity, being mindful of the tone of their speech,
and paying more attention to the content of what they said.” 33

Law enforcement technologies, specifically BWCs, can be ineffective without

sound policies governing their use. The ACLU argues that BWCs “can be a win-win—but
only if they are deployed within a framework of strong policies to ensure they protect the
public without becoming yet another system for routine surveillance of the public and
maintain public confidence in the integrity of those privacy protections. Without such a
framework, their accountability benefits would not exceed their privacy risks.” 34

D. LITERATURE REVIEW

Law enforcement’s use of surveillance technologies may create citizen fears of an

“Orwellian society.” Recent advances in technology become amplified, especially when
combined with outpaced legislation drawing additional concerns that law enforcement
agencies will have autonomy and misuse surveillance technologies. As such, in addition to
privacy concerns, public trust challenges of law enforcement exist. This review examines
the issues in two sections. The first section of this review provides an understanding of the
legal and ethical dilemmas and the onus placed on law enforcement to balance privacy and
security. The second section examines the literature related to public trust challenges faced
by law enforcement.

1. Privacy versus Security

Privacy concerns specific to the use of law enforcement technologies can be traced
back to the 1880s, when advancements in photography eased the use of equipment and
enabled portable cameras for use outside photo studios. The portable camera then began to
document the world in a new way and, as a result, created new privacy concerns. 35 An

33 Koen, Willis, and Mastrofski, 12.

34 Jay Stanley, Police Body-Mounted Cameras: With Right Policies in Place, a Win for All, version
2.0 (New York: American Civil Liberties Union, March 2015), 2, https://www.aclu.org/other/police-body-
mounted-cameras-right-policies-place-win-all.
35 Claudia Cuador, “From Street Photography to Face Recognition: Distinguishing between the Right
to Be Seen and the Right to Be Recognized,” Nova Law Review 41, no. 2 (2017): 5, https://nsuworks.nova.
edu/cgi/viewcontent.cgi?referer=https://scholar.google.com/&httpsredir=1&article=1998&context=nlr/.
11
evaluation of the literature of current frameworks, legal concerns, as well as law
enforcement and government recommendations for real-time FRT provides a deeper
understanding of the privacy challenges that the NYPD faces. The problem of how to
mitigate individual privacy concerns remains unanswered. No universal policies exist, and
therefore, no methods exist for data collection, sharing, and storage—raising ethical
concerns of how law enforcement organizations will use FRT. A New York University
study argues that law enforcement agencies should focus on ethical issues raised by sharing
data because expanding the use of new technologies compromises the desire of individuals
to have complete and absolute privacy. 36

Critics express concern over big data, massive storage, and the analytic capabilities
of new technologies. Ironically, citizens embrace certain technologies and fear others, even
though they offer similar benefits and challenges. Technologies like FRT experience a
combination of acceptance and rejection. For example, citizens laud FRT when used as a
security feature on a smartphone or to create a talking emoji, but in the hands of law
enforcement, that same technology creates negative feelings for many people. The reason
behind the negative association of FRT use centers on the potential for diminished levels
of privacy, which most citizens are accustomed to in a public setting. Privacy, as defined
by Ruth Gavison, is “a measure of the access others have to you through information,
attention, and physical proximity.” 37 Gavison’s definition contends that the lack of
“privacy” evokes a negative connotation and anything that alters total privacy is a violation,
an intrusion, or undesirable. Philosopher Jeffery Reiman provides a different definition of
privacy: “The condition under which other people are deprived of access to either some
information about you or some experience of you.” 38 Another legal scholar, Anita Allen,
contends that privacy has three dimensions: physical privacy, informational privacy, and
proprietary privacy. Informational privacy is arguably the leading concern among citizens

36 Julia Lane et al., Privacy, Big Data, and the Public Good: Frameworks for Engagement
(Cambridge: Cambridge University Press, 2014), 49, https://www.nyu.edu/projects/nissenbaum/papers/
BigDatasEndRun.pdf.
37 Helen Fay Nissenbaum, Privacy in Context: Technology, Policy, and the Integrity of Social Life
(Palo Alto, CA: Stanford University Press, 2010), 67.
38 Nissenbaum, 70.

12
when law enforcement uses surveillance technologies due to the potential reduction of
anonymity in public.

The public/private dichotomy infiltrates both legal and political discourse. The
point of contention is that the definitions of public and private vary depending on the
context. For example, the term private is associated with legal but personal engagements
in intimate settings or with an expectation of privacy when using the internet. The
public/private dichotomy also exists in government. In democratic societies, checks and
balances establish strict guidelines for intrusions into citizen privacy. Citizens have an
expectation of privacy, specifically when it involves government use of surveillance
technologies to track them. Some view these checks and balances as the right to restrict the
government access to personal records. Ironically, the U.S. Constitution does not explicitly
define privacy.

According to a 1990 survey, 79 percent of Americans claim that if the Declaration

of Independence was rewritten, they would be willing to add “privacy” in addition to “life,
liberty, and the pursuit of happiness.” 39 In another survey conducted in 2003, 95 percent
of citizens reported a concern with the internet as well as how personal information is
collected. 40 Actions, however, speak louder than words. Although many of the surveys
identified serious privacy concerns over the use of personal information, surveys
evaluating public benefits such as E-Z Pass, traceable search engines, discount shoppers’
cards, and other modes of convenience were deemed favorable. 41 Also, only 20 percent of
citizens claim to read privacy policies “most of the time,” and even less (7 percent)
complain about company privacy policies. 42

Potential community concerns associated with the NYPD’s use of real-time FRT
are relevant to the discussion. Public mistrust of the police still exists, both with misuse of
surveillance technologies and the methods by which law enforcement collects, shares, and

39 Nissenbaum, 103.
40 Nissenbaum, 103.
41 Nissenbaum, 103.
42 Nissenbaum, 103.

13
stores data. If the NYPD were to consider implementing real-time FRT, additional public
trust challenges might arise. Law enforcement’s deployment of real-time facial recognition
may prompt citizen resistance. A recent analysis by Georgetown Law argues that although
there are many benefits to FRT, significant risks remain to endanger individual privacy,
civil liberties, and civil rights. 43 Critics argue that without regulation, law enforcement
agencies would be free to collect and share information obtained from facial recognition
databases without oversight. An analysis by the Center on Privacy and Technology at
Georgetown Law argues that real-time FRT will create new paradigms in both policing and
individual conceptualizations of the definition of freedom. 44

Advocacy groups have resisted other law enforcement technologies, such as LPRs,
which in the past have created animosity between citizens and the police. The ACLU
argues that police organizations will use LPR technology as a tracking tool and that few
guidelines exist to maintain the integrity of the system. 45 In addition, the ACLU cites
specific cases where police officers misused LPR technology. 46 The vast majority of
individuals with knowledge of LPR technology view it as an efficient, non-evasive crime-
fighting tool. A report by RAND Corporation examines the benefits and challenges of LPR
technology, noting the ACLU’s belief that national use of LPR technology by law
enforcement becomes a gateway for other technologies to infringe on constitutional

43 Clare Garvie, Alvaro Bedoya, and Johnathan Frankle, The Perpetual Line-Up: Unregulated Police
Face Recognition in America (Washington, DC: Center on Privacy and Technology, Georgetown Law,
October 18, 2016), 1, https://www.perpetuallineup.org.
44 Garvie, Bedoya, and Frankle, 22.
45 Catherine Crump, You Are Being Tracked: How License Plate Readers Are Being Used to Record
Americans’ Movements (New York: American Civil Liberties Union, July 2013), 2, https://www.aclu.org/
issues/privacy-technology/location-tracking/you-are-being-tracked.
46 Mariko Hirose, “Documents Uncover NYPD’s Vast License Plate Reader Database,” American
Civil Liberties Union, January 25, 2016, https://www.aclu.org/blog/privacy-technology/location-tracking/
documents-uncover-nypds-vast-license-plate-reader-database.
14
rights. 47 An opposing article suggests the ACLU has made several misleading statements
about LPR technology, its use, and dissemination of collected data by law enforcement. 48

Among advocacy groups, an overwhelming concern remains that the government

is not willing to compromise security for privacy. Article 17 of the International Covenant
on Civil and Political Rights declares, “No one shall be subjected to arbitrary or unlawful
interference with his privacy, family, home or correspondence, nor to unlawful attacks on
his honour and reputation.” 49 The House Committee on Oversight and Government reform
oversaw a bipartisan hearing on March 22, 2017, to address law enforcement’s use of
FRT. 50 There was a strong emphasis on the policies created to mitigate privacy concerns.
Both subcommittees identified numerous issues, but none of those have spurred the
creation of new laws for facial recognition. While some political leaders and activist groups
agree there is a need for legislation on FRT, they do not agree on how to accomplish a
universal framework for law enforcement to follow. 51 The ACLU has provided
recommendations to the National Telecommunications and Information Administration for
an ethical framework for FRT, but specific legislation has yet to be created. 52

47 Keith Gierlack et al., License Plate Readers for Law Enforcement: Opportunities and Obstacles
(Santa Monica, CA: RAND Corporation, 2014), 17–18, https://www.rand.org/pubs/research_reports/
RR467.html.
48 Joel Griffin, “LPR Technology Comes of Age,” Security Info Watch, August 4, 2015, http://www.
securityinfowatch.com/article/12099279/lpr-technology-comes-of-age.
49 Francesca Bignami and Giorgio Resta, Human Rights Extraterritoriality: The Right to Privacy and
National Security Surveillance (Rochester, NY: Social Science Research Network, 2018), 1,
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3043771.
50 Hearing before the Committee to Review Law Enforcement’s Policies on Facial Recognition
Technology, House, 115th Cong., 1st sess., March 22, 2017, https://oversight.house.gov/hearing/law-
enforcements-use-facial-recognition-technology.
51 John Byrne, “Emanuel Raises Facial Recognition Tech in Taxi vs. Ride-Share Debate,” Chicago
Tribune, November 7, 2017, http://www.chicagotribune.com/news/local/politics/ct-met-rahm-emanuel-
rideshare-report-20171107-story.html.
52 Catherine Crump, “An Ethical Framework for Face Recognition,” American Civil Liberties Union,
July 16, 2013, https://www.ntia.doc.gov/files/ntia/publications/aclu_an_ethical_framework_for_face_
recognition.pdf.
15
 The vast majority of literature on automatic or real-time FRT suggests that
significant benefits exist for law enforcement agencies and corporations. 53 In addition, the
literature explains that the use of real-time FRT by law enforcement has practical benefits
because of recent technological developments. 54 The literature on the subject suggests that
data collection from FRT has favorable results when used in the field by law enforcement
and on the battlefield by the U.S. military. 55 Federal law enforcement agencies collectively
agree that transparency needs to exist between law enforcement and citizens when it
involves how law enforcement organizations collect and use biometric information. 56 The
Department of Homeland Security (DHS) took several steps to create policies and
procedures for biometric technologies that have strict standard operating procedures. DHS
uses the Automated Biometric Identification System for storing and processing biometric
information. 57 Privacy impact assessments provide citizens with detailed information on
how federal law enforcement collects, shares, and uses biometric information. 58

Some of the literature examines the success of real-time FRT in companies such as
Facebook and JetBlue. 59 Concerns still exist, however, in such areas as performance,

53 Future of Privacy Forum, Privacy Principles for Facial Recognition Technology (Washington, DC:
Future of Privacy Forum, December 2015), https://fpf.org/wp-content/uploads/2015/12/Dec9Working-
Paper-FacialRecognitionPrivacyPrinciples-For-Web.pdf.
54 Jeerawat Detsing and Mahasak Ketcham, “Detection and Facial Recognition for Investigation,” in
International Conference on Digital Arts, Media and Technology (Piscataway, NJ: IEEE, 2017), 407–409,
https://doi.org/10.1109/ICDAMT.2017.7905002.
55 Glenn J. Voelz, Rise of iWar: Identity, Information, and the Individualization of Modern Warfare
(New York: Skyhorse Publishing, 2018), 1–3, 10, https://ssi.armywarcollege.edu/pubs/display.cfm?
pubID=1292
56 Kimberly J. Del Greco, “Law Enforcement’s Use of Facial Recognition Technology,” Federal
Bureau of Investigations, March 22, 2017, https://www.fbi.gov/news/testimony/law-enforcements-use-of-
facial-recognition-technology.
57 “Privacy Impact Assessment for the Automated Biometric Identification System (IDENT),”
Department of Homeland Security, accessed August 8, 2018, 2–4, https://www.dhs.gov/sites/default/files/
publications/privacy-pia-nppd-identappendices-august2017.pdf; and W. Jeberson and Lucky Sharma,
“Survey on Big Data for Counter Terrorism,” International Journal of Innovations and Advancement in
Computer Science 4 (May 2015).
58 Department of Homeland Security, 5.
59 Wendy Davis, “Facial Recognition Technology Nab Criminals-and Raises Privacy Concerns,”
ABA Journal, October 2017, http://www.abajournal.com/magazine/article/
facial_recognition_technology_crime_ privacy.
16
intentional circumvention, and overall accuracy, but some are beginning to accept the
presence of FRT as part of everyday life. 60 A December 2017 report by the Center on
Privacy and Technology at Georgetown Law argues strongly against using FRT to identify
passengers. The report suggests, “DHS should justify its investment in face scans by
supplying evidence of the problem it purportedly solves.” 61 Furthermore, DHS needs to
conduct a cost analysis of the capabilities of the FRT air exit system to determine whether
it is beneficial and requires additional funding or it should be disbanded and release
finances to other initiatives. 62

The Constitution Project, a bipartisan committee, is attempting to lobby for strong

laws to protect citizens from the pervasive use of facial recognition. Jake Laperruque,
senior counsel at the Constitution Project, claims that facial recognition is an invasive
technology, and, if no laws exist, law enforcement agencies can and probably will abuse
it. 63 A report by the Constitution Project acknowledges that real-time facial recognition
can scan faces all day without limitations and without depleting police resources. 64 The
upshot, however, is that any individual can be scanned and possibly tracked by law
enforcement at events, demonstrations, political rallies, or protests—likely without any
legal authority. 65

Both advocacy groups and law enforcement desire a framework for how FRT,
particularly real-time facial recognition, should be implemented to ensure that law

60 Koichi Ito and Takafumi Aoki, “Recent Advances in Biometric Recognition,” ITE Transactions on
Media Technology and Applications 6, no. 1 (2018): 64–65, https://doi.org/10.3169/mta.6.64.
61 Laura M. Moy, Harrison Rudolph, and Alvaro M. Bedoya, Not Ready for Takeoff: Face Scans at
Airport Departure Gates (Washington, DC: Center on Privacy and Technology, Georgetown Law,
December 21, 2017), 3–4, https://www.airportfacescans.com/sites/default/files/Biometrics_Report__
Not_Ready_For_Takeoff.pdf.
62 Moy, Rudolph, and Bedoya, 5.
63 Jake Laperruque, “Preserving the Right to Obscurity in the Age of Facial Recognition,” Century
Foundation, October 20, 2017, https://tcf.org/content/report/preserving-right-obscurity-age-facial-
recognition/.
64 “Law Enforcement Facial Recognition Is a Powerful Surveillance Technology in Need of
Independent Checks and Limits,” Constitution Project, March 30, 2017, 1–2, https://constitutionproject.org/
wp-content/uploads/2017/03/Facial-Recognition-Statement-for-Record_The-Constitution-Project.pdf.
65 Constitution Project, 3.

17
enforcement officers do not infringe on individual constitutional rights. MIT researcher
Joy Buolamwini argues that FRT already infringes on constitutional rights because flaws
in facial recognition algorithms result in a predetermined bias—by design—that identifies
a person. 66 Problematically, under current law, any police organization can access and
disseminate facial recognition information on law-abiding individuals. In a 2016 article,
Laura Sydell notes that the Federal Bureau of Investigation (FBI)’s facial recognition
database includes driver’s license photos from 16 different states. 67 A general concern is
that this and similar databases will continue to grow and be shared with federal, state, and
local law enforcement agencies.

The FBI believes that law enforcement should move forward with some form of
“automated” FRT, with its analysis appearing to be in stark contrast to Sydell’s findings,
along with the findings of others. 68 FBI documents were the only sources in the literature
that focused on adversarial threats as the top priority—whereas other authors concentrated
their analyses on identifying privacy concerns. Although very little scholarly literature
exists on real-time facial recognition for use by law enforcement in urban subway systems,
there is, however, a recent article by Noah McClain that examines surveillance
technologies in the New York City subway system. He argues the importance of “looking
beyond the claims of technical efficacy in the study of security and surveillance to discover
how technologies of inspection and control work, as a means to cut through the heavy
rhetorical packaging in which they are sold to their publics.” 69 McClain’s position
forecasts the belief that FRT has arrived, whether the public approves of its use or not.

66 Michael Skirpan and Tom Yeh, “Designing a Moral Compass for the Future of Computer Vision
using Speculative Analysis,” in 2017 IEEE Conference on Computer Vision and Pattern Recognition
Workshops (Piscataway, NJ: IEEE, 2017), 64, http://openaccess.thecvf.com/content_cvpr_2017_
workshops/w16/papers/Yeh_Designing_a_Moral_CVPR_2017_paper.pdf.
67 Laura Sydell, “It Ain’t Me, Babe: Researchers Find Flaws in Police Facial Recognition
Technology,” NPR, October 25, 2016, http://www.npr.org/sections/alltechconsidered/2016/10/25/
499176469/it-aint-me-babe-researchers-find-flaws-in-police-facial-recognition.
68 Del Greco, “Law Enforcement’s Use of Facial Recognition.”
69 Noah McClain, “The Horizons of Technological Control: Automated Surveillance in the New York
Subway,” Information, Communication & Society 21, no. 1 (2018): 46, https://doi.org/10.1080/1369118X.
2016.1260624.
18
 A delicate balance exists between the citizens’ desire for privacy and their need for
security. These concerns often shift on a pendulum, swaying back and forth based on
national and global events. Such questions as “Is it even possible to maintain privacy in an
age where our daily actions can be monitored?” are reasonable and directly correlate with
advancements in technology and new social media platforms. 70 FRT has a place in the
natural progression of technological tools—from the installation of police call boxes in
1877, to fingerprint analysis, to CCTV cameras, and finally to more modern law
enforcement technologies such as BWCs and drones.

2. Transparency

The foundation of policing is based on collaboration with citizens, public support,

police transparency, and citizen acceptance of policing strategies, tactics, and policies.
Lewis, Daniel, and Kalalea contend that in every police–citizen interaction, police earn a
new friend or a new enemy. 71 This reality illustrates the dire need for police officers to
create as many positive interactions as possible with citizens. Historically, efforts have
been made to involve the community in policing. Most researchers contend that community
policing encompasses three main ideas:

1) collaborative partnerships between the community and the police, while

collectively address and prioritize community concerns; 2) organizational
transformation, through which management and personnel develop the
structure to collaborate with the community effectively; and 3) problem
solving through “proactive and systematic examination of identified
problems to develop and evaluate effective responses.” 72

70 Susan McCoy, “O’ Big Brother Where Art Thou?: The Constitutional Use of Facial-Recognition
Technology,” John Marshall Journal of Computer & Information Law 20, no. 3 (Spring 2002): 471.
71 Stephanie Daniel, Nicole Lewis, and Kalalea Kalalea, “In the Wake of Broken Windows Policing
How Aggressive Policing Contributed to East Harlem Residents Distrust of Police” (master’s capstone,
City University of New York, 2016), 2.
72 Kathleen O’Reilly, “Transparency, Accountability, and Engagement: A Recipe for Building Trust
in Policing” (master’s thesis, Naval Postgraduate School, June 2017), 11.
19
In the NYPD, the first community policing strategy began in 1984 as a pilot program in
the 72nd Precinct. 73 The NYPD rolled out the community patrol officer program (CPOP)
during an era when significant mistrust of the police existed. The NYPD designed a
strategy to combat community mistrust, instructing the community policing officer to
foster relationships with community members and to solve the issues of the community
collectively. The program, however, never gained acceptance from within the NYPD. One
researcher acknowledges that other officers identified community police officers as
“privileged social workers, detached from the department’s daily crime-fighting
pressures.” 74 It was not until 2014, because of increased citizen concern, that the NYPD
looked to formulate a new plan to address citizen mistrust.

Citizens, legislators, and members of the media have scrutinized several recent fatal
police-involved use-of-force incidents. Former Secretary of State Hillary Clinton stated
during the 2016 presidential election, “The deaths of Alton and Philando drove home how
urgently we need to make reforms to policing and criminal justice . . . how we cannot rest
until we root out implicit bias and stop the killings of African-Americans.” 75 Clinton
eluded to separate incidents—the deaths of Alton Sterling and Philando Castile, who were
both shot and killed by police. Demonstrations, protests, and violence against the police,
including several assassinations of police officers, occurred around the country as
retribution for vigorously contested police-involved fatal use-of-force incidents.

The recent formation of new civil rights groups—Black Lives Matter, the Anti-
Police Terror Project, and Communities United for Police Reform (CUFPR) among
others—illustrates a growing citizen response to mistrust of law enforcement. CUFPR has
created a brochure to inform citizens of their rights following an interaction with the

73 Eli B. Silverman, NYPD Battles Crime: Innovative Strategies in Policing (Boston: Northeastern
University Press, 1999), 55.
74 Silverman, 56.
75 “Full Transcript of Hillary Clinton’s NAACP Speech: ‘This Madness Has to Stop,’” Fortune, July
18, 2016, http://fortune.com/2016/07/18/hillary-clinton-speech-naacp-transcript/.
20
police. 76 The brochure reminds citizens, “Police officers can be unpredictable and can
quickly escalate a situation, particularly if they feel unsafe, disrespected, or that they don’t
have control of the situation.” 77 The brochure’s message depicts inimical relations and
mistrust of police by citizens; it also illuminates the misnomers that reflect unintentional
misunderstandings and implicit bias of both police and citizens. Recent literature on police
relations with citizens focuses less on law enforcement technologies and more on police
use-of-force, implicit bias by police officers, and the need for transparency from law
enforcement.

In 2014, one of the priorities of the newly appointed Mayor of New York City, Bill
de Blasio, was to bridge the existing gap between citizens and the NYPD. Mayor de Blasio
immediately appointed William J. Bratton to lead the NYPD, and his role was clear:
continue to reduce crime and cultivate relationships between the police and the
communities they serve. Bratton’s appointment was his second tenure as police
commissioner of New York City—he was already familiar with its dynamics and
challenges. Bratton brought with him decades of police leadership from when he left the
NYPD in 1996, but by 2014, new challenges had emerged, particularly with damaged
relations between communities and the NYPD. In previous years, the NYPD’s stop,
question, and frisk (SQF) policy resulted in criticism nationwide and eventual federal
monitoring after several lawsuits involving unlawful stops. Deficiencies in both policy and
officer discretion caused critics to label “stop and frisk” a biased practice. 78 Statistical SQF
data and citizen perception accentuated the fact that the existing practice needed an
evaluation. In 2011, there were 685,724 persons stopped, and the stop did not result in an
arrest 88 percent of the time. 79 In 2013, a legislative decision by Judge Analisa Torres

76 “Know Your Rights! Help End Discriminatory, Abusive & Illegal Policing!” Communities United
for Police Reform, accessed May 6, 2018, http://changethenypd.org/resources/know-your-rights-help-end-
discriminatory-abusive-illegal-policing.
77 Communities United for Police Reform.
78 Michael D. White and Henry F. Fradella, Stop and Frisk: The Use and Abuse of a Controversial
Policing Tactic (New York University Press, 2016), 7.
79 Al Baker, “Street Stops by New York City Police Have Plummeted,” New York Times, May 30,
2017, https://www.nytimes.com/2017/05/30/nyregion/nypd-stop-and-frisk.html.
21
ruled the practice of SQF as unconstitutional. 80 Federal Judge Shira A. Scheindlin
appointed independent monitor Peter L. Zimroth to oversee changes in the SQF strategy.
In 2017, there were 11,553 people questioned under the practice. Zimroth stated, “Things
are trending in the right direction” with fewer stops on record than in prior years. 81 Even
with fewer stops of individuals, violent crimes continue to decline in New York City. In
2017, there were 292 homicides, the lowest on record since 1951 (243). 82 Recent crime
reductions can be contributed to improved transparency within the NYPD and the
implementation of neighborhood policing.

Neighborhood policing is an innovative strategy, dubbed a new way of policing,

developed by the NYPD to rebuild trust between the community and police. Neighborhood
policing does not derive from CPOP because its foundational principles and objectives are
different. Neighborhood coordination officers help rebuild trust and get to know the people
in the community. Moving forward, the NYPD has created a new strategy to rebrand its
image. In 2018, a new department initiative, called “Build the Block,” became part of the
neighborhood-policing model.

The strategy behind neighborhood policing is to facilitate collaboration among

members of the community to share the responsibility of safety in New York City
neighborhoods, allowing communities to work in partnership with the NYPD. Build the
Block neighborhood safety meetings “are strategy sessions between local police officers
and the public they serve. These meetings have two simple goals: identify the public safety
challenges of a specific neighborhood and discuss potential solutions.” 83 Transparency and
trust between community members and police are essential parts of the crime-fighting
mission. Initiating new strategies, technologies, or policies, increases the NYPD’s chances

80 Baker.
81 J. David Goodman, “Court-Appointed Police Monitor Has Fought for City and against It,” New
York Times, August 13, 2013, https://www.nytimes.com/2013/08/14/nyregion/court-appointed-police-
monitor-has-fought-for-city-and-against-it.html.
82 “Crime Statistics,” New York City Police Department, accessed July 17, 2018, https://www1.nyc.
gov/site/nypd/stats/crime-statistics/crime-statistics-landing.page.
83 “Build the Block,” New York City Police Department, accessed July 17, 2018, https://www1.nyc.
gov/site/nypd/bureaus/patrol/buildtheblock.page.
22
for overall success and of public acceptance because citizens feel involved and express
their concerns to the department.

E. RESEARCH DESIGN

This thesis uses the comparative case study method to examine the benefits and
challenges of LPR technology to address how the NYPD can mitigate concerns and
evaluate real-time FRT for use in the New York City subway system. The comparison
examines the NYPD’s use of LPR technology and then identifies each of the benefits and
challenges. The benefits were identified through the literature, public discourse, case law,
internal open-source NYPD documents, and studies from scholars and advocacy groups.
The research addresses the extent to which LPRs are analogous to FRT and how the use of
LPRs can provide the NYPD with a roadmap for public support of real-time FRT. The
possible vulnerabilities of real-time FRT may include reduced levels of citizen privacy,
public trust challenges, and overall unreliability. The benefits of LPR technology and FRT
examined in this thesis are divided into two categories. The first category illustrates the
crime prevention benefits of LPRs and FRT, and the second identifies law enforcement
and public benefits. The order of each categorical benefit begins with the most significant
use for each technology.

This thesis asserts that the NYPD does not have to start from scratch to figure out
how to mitigate the challenges to reap public benefits of real-time FRT. The NYPD and
other law enforcement organizations already use such surveillance technologies as LPRs,
which scan license plates to ascertain pertinent information about vehicles and their
registered owners. Many of the challenges associated with LPRs have already been
mitigated, resulting in citizen acceptance and expanded use. The public benefit and overall
citizen acceptance of LPR technology can provide a framework for the NYPD to
implement real-time FRT.

Advancements in FRT have created significant benefits for law enforcement

organizations to reduce crime and potentially mitigate acts of terrorism. Although
significant benefits exist, the NYPD faces challenges that require a thorough evaluation
before implementing real-time FRT for use in the New York City subway system. The

23
NYPD, however, can leverage lessons learned from LPR technology. Both LPR
technology and FRT are similar in that they share many of the same benefits and
challenges. Documents show that more law enforcement agencies in the United States are
implementing LPR technology. The NYPD uses LPRs, and recent technology
advancements have created additional benefits. Although several new benefits have been
identified, new challenges have emerged. The NYPD may need to re-evaluate privacy,
accuracy, and public trust concerns with LPR technology.

F. CHAPTER OUTLINE

An overview of LPRs, their use in law enforcement, and the identified benefits and
challenges of the technology are presented in Chapter II. Law enforcement use of FRT has
negative perceptions, and critics of the technology are quick to document concerns that are
unintentionally analogous to the concerns of LPRs. Chapter III provides an overview of
FRT, its current and projected use by law enforcement, and an examination of the
technology’s benefits and challenges. Chapter IV illustrates the ability for decision-makers
to mitigate the challenges of law enforcement’s use of real-time FRT. With public support,
the NYPD can effectively implement real-time FRT in the New York City subway system.
The chapter then concludes with the lessons learned from LPRs and FRT and provides
recommendations to the NYPD for next steps and further research.

24
 II. LICENSE PLATE READERS

Communities benefit from LPR technology despite privacy concerns and the
contention of critics—such as the ACLU, which argues that law enforcement agencies
should not be able to collect and store information on law-abiding citizens not suspected
of any criminal wrongdoing. 84 This chapter provides an overview of LPR technology,
including its definition and history as well as the benefits of LPR technology for law
enforcement and the public. It then examines the challenges of LPR. This chapter shows
that the benefits of LPR technology outweigh its challenges. The benefits and challenges
of LPR technology in this chapter provide a framework to analyze the benefits and
challenges of real-time FRT.

A. OVERVIEW

This section explains how LPR technology was developed, how it was first used,
and how it is used today. Then, it explains how officers use LPR technology, provides the
NYPD’s definition of LPR technology, and describes how LPRs supplement police
operations.

1. History of License Plate Reader Technology

In 1976, the Police Scientific Development Branch of the Home Office in the
United Kingdom invented LPR technology, also known as automatic number plate
recognition, which captured the image of a license plate and attempted to match it against
a pre-determined hot-list containing reported information on stolen vehicles. 85 Initially,
LPRs were developed to assist law enforcement in identifying stolen vehicles, but in April
1993, terrorists detonated a bomb inside a truck in the financial area of London. Following

84 Ben Eisler, “ACLU Concerned Automatic License Plate Readers May Invade Privacy,” WJLA
News, July 30 2012, http://wjla.com/news/local/aclu-concerned-red-light-cameras-may-invade-privacy-
78301.
85 David J. Roberts and Meghann Casanova, Automated License Plate Recognition Use by Law
Enforcement: Policy and Operational Guide, Summary, doc. 239605 (Alexandria, VA: International
Association of Chiefs of Police, September 2012), 3, https://www.ncjrs.gov/pdffiles1/nij/grants/239605.
pdf.
25
the attack, City of London Police included LPRs in its new strategy to secure London,
known today as the “ring of steel.” 86 The goal of the City of London Police was to apply
combat-style strategies in civilian settings to protect citizens, or “fortress urbanism.” 87
LPR technology gave police valuable intelligence by identifying every vehicle that entered
London—which was necessary because the Irish Republican Army had identified central
London as an ideal location to conduct bombings. 88 As acts of terrorism continued in the
1990s, the use of LPR technology expanded with little citizen resistance. In 1997, LPR
cameras were placed at the entrances of the ring of steel. In 2002, the use of LPRs expanded
to nine different police agencies. 89 In 2003, LPRs were used to facilitate traffic congestion
tolls, and in 2006, the data storage capabilities were enhanced for a capacity of 50 million
plate scans per day. 90

In 2008, the NYPD launched a similar strategy to strengthen resiliency capabilities

by using technology, increased police presence, and collaboration with the private sector
to protect lower Manhattan and sensitive locations, such as the Stock Exchange, the World
Trade Center, and Federal Plaza. This strategy, known as the Lower Manhattan Security
Initiative (LMSI), uses both mobile and stationary LPR cameras, uniformed police
presence, and technology to detect threats and conduct “pre-operational terrorist
surveillance,” which augments the deployment of specially trained police officers and
integrated technology platforms. 91 Since 2008, the LMSI has expanded, involving the
better integration of technologies such as radiation pagers, LPR technology, and CCTV
cameras. The success of the LMSI has warranted further expansion into other areas of

86 J. Mullins, “Ring of Steel II: New York City Gets Set to Replicate London’s High-Security Zone,”
IEEE Spectrum 43, no. 7 (July 2006): 12, https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=
1652996.
87 “Ring of Steel,” MAS Context, June 25, 2014, http://www.mascontext.com/issues/22-surveillance-
summer-14/ring-of-steel/.
88 MAS Context.
89 “Protection of Freedoms Bill,” UK Parliament, accessed August 8, 2008, https://publications.
parliament.uk/pa/cm201011/cmpublic/protection/memo/pf11.htm.
90 UK Parliament.
91 “Counterterrorism,” New York City Police Department, accessed May 14, 2018, https://www1.nyc.
gov/site/nypd/bureaus/investigative/counterterrorism.page.
26
Manhattan. 92 In addition to the NYPD, 71 percent of law enforcement agencies in the
United States were using LPR technology as of 2012. 93 Five years later, in 2017,
researchers projected that approximately 85 percent of police agencies would obtain
funding to purchase LPR equipment. 94

In 2013, the NYPD expanded the use of LPR readers throughout New York City. 95
Citing that no laws prevent the use of surveillance technology to prevent crime, the NYPD
expanded LPR use and leveraged its ability to perform additional functions. For example,
in addition to identifying stolen vehicles, the NYPD often examines LPR data to assist in
police investigations. Officers are instantly able to recover a stolen vehicle based on real-
time information supplied to an officer on patrol. Furthermore, detectives can leverage LPR
data to piece together on-going investigations. With new advancements in LPR technology,
local governments, law enforcement, and private companies all leverage LPRs to perform
various tasks—real-time traffic monitoring, toll collections, parking-lot access control, and
traffic enforcement—not to mention cameras to enforce traffic laws. 96

Since 2016, the public has accepted the NYPD’s use of LPR technology due to
tangible public benefits. An example of such benefits occurred in June 2018, when a
teenage boy was attacked on a Bronx street corner by several perpetrators. In a case of
mistaken identity, gang members stabbed the boy to death, but his murder was captured on
surveillance cameras and cell phone videos. When the videos were posted on social media,
his horrific and untimely death created outrage among both citizens and law enforcement.

92 New York City Police Department, “Midtown Manhattan Security Initiative” (press release, NYPD,
September 20, 2010), http://www.nyc.gov/html/nypd/html/pr/pr_2010_midtown_security_initiative.shtml.
93 Chuck Wexler, “Introduction,” in How are Innovations in Technology Transforming Policing
(Washington, DC: Police Executive Research Forum, January 2012), iii, http://www.policeforum.org/
assets/docs/Critical_Issues_Series/
how%20are%20innovations%20in%20technology%20transforming%20policing%202012.pdf.
94 Gierlack et al., License Plate Readers for Law Enforcement, 8.
95 Chris Francescani, “NYPD Expands Surveillance Net to Fight Crime as well as Terrorism,”
Reuters, June 21, 2013, https://www.reuters.com/article/usa-ny-surveillance/nypd-expands-surveillance-
net-to-fight-crime-as-well-as-terrorism-idUSL2N0EV0D220130621.
96 Shan Du et al., “Automatic License Plate Recognition: A State-of-the-Art Review,” IEEE
Transactions on Circuits and Systems for Video Technology 23, no. 2 (February 2013): 311–312,
https://doi.org/10.1109/TCSVT.2012.2203741.
27
Numerous gang members had been involved and were considered an extreme risk to citizen
safety. After an extensive investigation, NYPD detectives arrested most of the perpetrators
within a few days. An LPR caught the last remaining suspect on the loose along Interstate
84 in Danbury, Connecticut. 97 High-profile cases in which dangerous perpetrators are
apprehended because of LPRs demonstrate the technology’s benefit to the safety of
citizens.

2. How License Plate Readers Work

LPR technology operates using high-speed infrared cameras that first photograph
license plates, then create an image of the plate, and match the image to a law enforcement
hot-list, which can compare the scanned image with local, state, and federal databases. The
NYPD clearly defines an LPR and in its policy requires that an LPR operator verify
information received from a device before taking police action:

An LPR device can identify a target plate within seconds. LPR devices may
read each license plate passed and alert the LPR operator when there is a
match to a list of specific or partial license plate numbers. The LPR device
is not automatically connected to NYSPIN [the New York State Police
Identification Network], is not programmed to scan the state of registration
on a license plate and will activate upon a partial scan match. Therefore, it
is absolutely essential that the LPR operator verify the current status of a
vehicle through NYSPIN prior to initiating any law enforcement action
(e.g., arrest, summons, etc.). 98

The verification of LPR information assists the officer in determining whether to conduct
a known-risk vehicle stop (KRVS), a traffic stop that potentially has increased risks for the
police officers initiating it. An officer conducting a KRVS must exercise additional tactical
considerations and make a more cautious approach, based on the potential threat identified
by the LPR. Figure 1 illustrates how law enforcement uses LPR technology.

97 Jim Shay, “License Plate Reader Help Nab Suspect in Slaying of NYPD Explorer,” Officer, July
11, 2018, https://www.officer.com/command-hq/technology/traffic/lpr-license-plate-recognition/news/
21012850/license-plate-reader-help-nab-suspect-in-slaying-of-nypd-explorer-lesandro-junior-guzmanfeliz.
98 New York City Police Department, Patrol Guide: “Padlock Law” Program, 291–31 (New York:
NYPD, July 1, 2014), https://www1.nyc.gov/assets/nypd/downloads/pdf/public_information/public-
pguide3.pdf.
28
 Figure 1. Notional Large-Scale, Integrated Departmental LPR Setup 99

It would be impossible for officers to conduct license plate checks at the same rate
as an LPR reader, which can analyze up to 1,800 license plates per minute. Each LPR
device costs between $10,000 and $22,000, depending on the manufacturer and the type—
vehicle-equipped, portable, or fixed. 100 Despite the cost, differing LPR deployments allow
law enforcement to apply various crime-fighting or tactical strategies. For example, LPRs
installed at fixed locations, such as bridges and tunnels, provide law enforcement specific
information regarding individual vehicle whereabouts and the ability to identify all vehicles
entering and exiting specific locations within New York City. The LPR cameras installed
on patrol vehicles can create a record of license plates near a crime scene, which may yield
valuable evidence in future criminal proceedings. For example, law enforcement officers
can examine LPR data near a crime scene and use them to supplement other pieces of

99 Source: Gierlack et al., License Plate Readers for Law Enforcement, 65.
100 Roberts and Casanova, Automated License Plate Recognition, 2.

29
evidence to identify potential suspects or witnesses. Portable LPR cameras provide law
enforcement agencies the versatility to record license plate data in locations where fixed
LPR cameras do not exist, but the surveillance needs—at massive demonstrations, parades,
or events—may be necessary.

B. BENEFITS OF LPR TECHNOLOGY

Due to recent improvements, law enforcement agencies have been researching new
ways to use LPR technology, expanding its uses exponentially from its original purpose to
identify stolen vehicles. Over the last 22 years, the number of stolen vehicles has drastically
decreased, and although no specific evidence connects the decline to LPR technology, such
a trend implies a nexus between technology and efficiencies in policing. As a matter of
course, other law enforcement strategies and technologies, including better vehicle security
features and GPS systems, have also aided this trend in terms of prevention. Nonetheless,
cars are still stolen as thieves adapt to theft deterrence technologies, which suggest a
continued need for LPR technology. In 1996, nearly 1.4 million vehicles were reported
stolen in the United States while, in 2015, only 723,186 vehicles were stolen—a 53 percent
decrease, despite an increase of 56,190,248 people and 58 million more registered vehicles
nationwide. 101 While identifying or tracking stolen vehicles remains the primary use of
LPRs in law enforcement, other public safety benefits exist. LPRs can assist in decreasing
crime, potentially prevent and help investigate acts of terrorism, identify wanted persons,
assist in detective investigations, provide additional officer safety, support individuals with
mental health needs, locate missing persons, and be used in conjunction with other law
enforcement technologies such as facial recognition.

1. Crime Prevention Benefits

The most important benefits of LPRs all involve preventing or reducing crime. The
use of LPRs has tangible benefits in reducing and investigating crimes.

101 “2015 Crime in the United States,” Federal Bureau of Investigation, accessed June 3, 2018,
https://ucr.fbi.gov/crime-in-the-u.s/2015/crime-in-the-u.s.-2015/tables/table-1; and “Number of Motor
Vehicles Registered in the United States from 1990 to 2016 (in 1,000s),” Statista, accessed June 3, 2018,
https://www.statista.com/statistics/183505/number-of-vehicles-in-the-united-states-since-1990/.
30
 a. Preventing Violent Crime

With a comprehensive policy and crime reduction strategy in place, in addition to

recovering stolen vehicles, LPRs can assist police departments in reducing violent crimes
such as homicide, rape, robbery, and felony assault. In 2004, the Ohio State Highway Patrol
conducted a four-month test of LPR technology to determine its effectiveness in crime
fighting. Officers made 23 arrests and recovered 24 stolen vehicles. 102 Since the
implementation of LPR technology by the Rockford, Illinois, Police Department, there has
been an 8 percent reduction in violent crime, and property crime has decreased by 5
percent. 103 Koper, Taylor, and Woods contend in a 2013 study that LPR technology could
be an effective tool in reducing crime in targeted areas. 104 By analyzing crime data,
officers can deploy LPRs in specific locations with increases in crime to identify and arrest
operators of stolen vehicles and investigate vehicles of past violent crimes—such as
robberies and assaults—to identify potential patterns. In addition, law enforcement officers
can use LPRs to investigate non-violent crimes, such as thefts from vehicles, to identify
similarities in specific areas.

b. Investigating Terrorism

LPR technology can be used to help investigate and may prevent terror attacks.
Terrorists are increasing their use of vehicles to conduct vehicle-ramming attacks (VRAs)
against innocent pedestrians. More than a dozen VRAs have occurred around the world
since 2014, claiming the lives of more than 170 people. 105 The NYPD used LPRs to
investigate a recent VRA in lower Manhattan. On October 31, 2017, Sayfullo Saipov drove

102 Tim Dees, “Finding Stolen Vehicles,” Hendon Media Group, accessed June 3, 2018,
http://www.hendonpub.com/resources/article_archive/results/details?id=3901.
103 Jeff Kolkey, “How One Illinois City Uses Automatic License Plate Readers and Other Police
Tech,” Government Technology, May 21, 2018, http://www.govtech.com/dc/How-One-Illinois-City-Uses-
Automatic-License-Plate-Readers-and-Other-Police-Tech.html.
104 Christopher S. Koper, Bruce G. Taylor, and Daniel J. Woods, “A Randomized Test of Initial and
Residual Deterrence from Directed Patrols and Use of License Plate Readers at Crime Hot Spots,” Journal
of Experimental Criminology 9, no. 2 (June 2013): 217, 213–244, https://www.voiceofsandiego.org/wp-
content/uploads/2018/04/Koper-et-al-2.pdf.
105 “Vehicles as Weapons of Terror,” Counter Extremism Project, May 22, 2017, https://www.
counterextremism.com/vehicles-as-weapons-of-terror.
31
a truck onto a pedestrian jogging path in lower Manhattan for approximately one mile,
killing eight people and injuring 11. 106 Responding officers shot and wounded Saipov
before placing him under arrest. Crime scene investigators recovered documents inside the
truck, which contained an Islamic State flag and documentation of his radicalization.

According to John Miller, NYPD deputy commissioner of counterterrorism and

intelligence, LPR data tracked Saipov’s movements leading up to the attack. When
examined, his movements offered additional investigative value. Detectives were able to
track Saipov’s movements throughout the City and combine video footage from CCTV
cameras along his traveled route. In addition, detectives analyzed LPR data to identify
potential accomplices and examine Saipov’s four prior trips into Manhattan. Furthermore,
the NYPD maintains close relationships with members of various truck rental businesses
to prevent VRAs or truck bombing attacks. Starting in 2016, the NYPD began visiting
truck-rental companies and discussed suspicious indicators and methods to report
suspicious activity. Any rented vehicle that is later deemed suspicious, based on an
investigation by law enforcement, can be entered into an LPR database and stopped by law
enforcement, which may mitigate a future attack.

The NYPD does not release the specific number of LPR cameras in New York City,
but current cameras do provide near-instant information that can be accessed using the
department’s domain awareness system (DAS), which links department databases into one
searchable application. LPR data was an essential part of the analysis to retrace the
aforementioned attack in New York City. Also, law enforcement can use data obtained
from LPRs to produce valuable information for future movements of potential terrorists
moving through the streets of New York City, which may shed some insight into future
target selection. In investigating the 2017 VRA, LPR data enabled the NYPD to reconstruct
a timeline of the events that preceded the attack. In addition, as LPR use by different law
enforcement agencies expands, traditional and new uses will benefit law enforcement
crime-fighting and counter-terrorism strategies.

106 Corey Kilgannon and Joseph Goldstein, “Sayfullo Saipov, the Suspect in the New York Terror
Attack, and His Past,” New York Times, November 1, 2017, https://www.nytimes.com/2017/10/31/
nyregion/sayfullo-saipov-manhattan-truck-attack.html.
32
 c. Real-Time Identification of Wanted Persons and Known Suspects

LPR real-time data, obtained from mobile, stationary, and portable readers, can
provide license plate information that may become valuable in other active criminal
investigations. 107 Larger police departments, such as the NYPD, have numerous LPR
cameras in mobile and fixed positions throughout their cities. Law enforcement can, as
needed, search the captured real-time data for valuable investigatory information. For
example, when television news reporter Vester Lee Flanagan was murdered in Virginia on
live television, law enforcement relied on real-time LPR data to apprehend his former
colleague for the crime. 108 Law enforcement officers added the suspect’s plate number to
the police hot-list, which led to his arrest.

d. Investigative Assistance

NYPD software has the ability to perform data integration, combining

information—complaint reports, summons activity, arrest activity, 3–1-1 and 9–1-1 calls,
wanted persons, police department vehicle locations, LPR data, and other information—to
portray a detailed snapshot of the current environment in New York City. The NYPD uses
its DAS to capture and create a tactical response based on real-time data. As a result, the
NYPD can swiftly make accurate assessments of a vehicle’s present and past locations, as
well as other metrics, in real time. 109

This saturation of data can be useful in a variety of ways. For example, following a
comprehensive criminal investigation, detectives used LPR technology to obtain probable
cause in the arrest of Marat G. Mikhaylich for a string of nine bank robberies. 110
Mikhaylich managed to elude capture, but after robbing a bank in Edison, New Jersey, he

107 Laura J. Moriarty, ed., Criminal Justice Technology in the 21st Century, 3rd ed. (Springfield, IL:
Charles C. Thomas Publisher, 2017), 263. Smaller law enforcement organizations, however, may be limited
in effectiveness due to a lack of real-time data, mainly because of the insufficient number of LPRs.
108 Nathan Tempey, “The NYPD Is Tracking Drivers across the Country Using License Plate
Readers,” Gothamist, January 26, 2016, http://gothamist.com/2016/01/26/license_plate_readers_nypd.php.
109 Jeberson and Sharma, “Survey on Big Data,” 202.
110 Al Baker, “License Plate Cameras Aid in Police Investigations,” New York Times, April 11, 2011,
https://www.nytimes.com/2011/04/12/nyregion/12plates.html.
33
stole a livery cab and drove to Queens, New York. LPR scans identified the stolen cab, and
when detectives located it the next morning, they were able to arrest Mikhaylich inside the
same vehicle. 111 His arrest was the result of the LPR technology’s ability to analyze
historical LPR data. Without it, Mikhaylich would have remained on the streets,
committing additional violent crimes.

2. Law Enforcement and Public Benefits

In addition to crime prevention benefits, LPRs may increase officer safety and assist
officers in identifying drivers with documented medical conditions who may have driving
restrictions, as well as locating missing or abducted persons.

a. Officer Safety

There are three benefits of using LPRs that contribute to officer safety. The first
benefit increases fairness and officer safety in policing practices because LPRs assist
officers in facilitating non-biased policing. LPRs cannot choose which plates they will
scan, and the results are based on information obtained from local, state, and federal
databases. Moreover, the technology may reduce racial profiling, which also helps to build
trust between police and communities and, thus, creates a safer environment for
officers. 112 According to a report by RAND Corporation, LPRs scan every plate and
cannot evaluate anything about the occupants of the vehicle. 113

The second benefit of LPR technology contributes to officer safety as it identifies

stolen vehicles or registered owners wanted in connection with past crimes. Such
information identified through a vehicle hot-list gives officers a heightened sense of
awareness when approaching a vehicle, so officers approach a vehicle with extra caution
or request the assistance of additional officers to investigate the vehicle and occupants in
question.

111 Baker.
112 Gierlack et al., License Plate Readers for Law Enforcement, 14.
113 Gierlack et al., 14.

34
 Finally, LPR technology helps officers on patrol concentrate on other tasks, such
as driving their patrol cars safely and performing other functions that require an officer to
stay alert for potential hazards. 114 In 2017, traffic collisions were the leading cause of
deaths, with 47 police officers killed in the line of duty. 115 LPR technology scans vehicle
plates, enabling the officer to keep focused on driving.

b. Mental Health Support

Approximately 9.8 million Americans—one out of every 25—experience serious

mental-health issues. 116 The significant, sensitive needs of approximately 4 percent of the
U.S. population are now better addressed by police departments. In addition to identifying
drivers operating vehicles without valid insurance, LPRs can detect registered vehicle
owners having suspended or revoked licenses. On March 5, 2018, Dorothy Bruns suffered
a seizure while driving in the middle of the day on a Brooklyn street and tragically killed
two young children. Officers arrested the vehicle operator and, after a lengthy
investigation, charged her with manslaughter, based on information that she had suffered
a seizure two months earlier while driving and had been instructed by doctors not to drive
for one year. 117 To prevent future tragedies involving drivers operating motor vehicles
with known medical conditions that would preclude them from driving, New York City
Mayor Bill de Blasio pushed for legislation mandating that doctors notify the Department
of Motor Vehicles of drivers with known conditions that could cause them to lose
consciousness. 118 LPRs can identify owners operating their vehicles with suspended

114 Moriarty, Criminal Justice Technology, 264–265.

115 National Law Enforcement Officers Memorial Fund, 2017 End of Year Officer Fatalities Report
(Washington, DC: NLEOMF, 2017), http://www.nleomf.org/assets/pdfs/reports/fatality-reports/2017/2017-
End-of-Year-Officer-Fatalities-Report_FINAL.pdf.
116 “Mental Health by the Numbers,” National Alliance on Mental Illness, accessed June 17, 2018,
https://www.nami.org/Learn-More/Mental-Health-By-the-Numbers.
117 Colin Moynihan, “Driver Charged with Manslaughter in Deaths of 2 Children,” New York Times,
May 3, 2018, https://www.nytimes.com/2018/05/03/nyregion/driver-manslaughter-brooklyn-children.html.
118 “Vision Zero: Mayor de Blasio Pushes for Tougher State Laws to Keep Dangerous Drivers off the
Streets,” City of New York, March 15, 2018, https://www1.nyc.gov/office-of-the-mayor/news/136-18/
vision-zero-mayor-de-blasio-pushes-tougher-state-laws-keep-dangerous-drivers-off-streets#/0.
35
licenses, and therefore, officers can proactively stop and make a summary arrest, as long
as the registered owner is also the operator of the vehicle.

In a 2013 experiment, Lum, Hibdon, Cave, Koper, and Merola evaluated the
potential use of LPR technology to identify vehicles being operated by owners with
suspended licenses. 119 The evaluation determined that although LPR technology does not
act as a deterrent to operate a motor vehicle with a suspended license, it could identify
registered owners of vehicles who have suspended licenses and could be stopped and
possibly arrested. 120 LPR technology, in conjunction with a collaborative effort with the
New York State Department of Motor Vehicles, legislators, and the district attorney’s
offices, can be leveraged to mitigate similar circumstances in the future. Data that identify
individuals with suspended or revoked licenses can be entered into department LPR hot-
lists to alert police officers in the field. Officers are legally permitted to stop a vehicle to
verify a driver’s credentials before taking enforcement action, based on LPR data that
identify possible violations or crimes relating to a vehicle or its registered owner.

c. Missing Persons

When law enforcement activates an AMBER alert for a missing child, law
enforcement has several methods to alert the public, including digital billboards and cell
phones. 121 In addition, if the alert involves a vehicle, law enforcement officers enter the
plate number into the LPR hot-list. The New York State Division of Criminal Justice
Services has created specific guidelines for dealing with AMBER alerts—specifically how
law enforcement should react to ensure the proper dissemination of intelligence. 122 In

119 Cynthia Lum et al., “License Plate Reader (LPR) Police Patrols in Crime Hot Spots: An
Experimental Evaluation in Two Adjacent Jurisdictions,” Journal of Experimental Criminology 7, no. 4
(2011): 321, https://www.voiceofsandiego.org/wp-content/uploads/2018/04/Lum-et-al-2.pdf.
120 Lum et al., 332.
121 The AMBER (America’s Missing: Broadcast Emergency Response) alert was created in 1996,
after nine-year-old Amber Hagerman was kidnapped and tragically murdered.
122 New York State Division of Criminal Justice Services, Operation of License Plate Readers in New
York State: Suggested Guidelines (Albany: NYS DCJS, June 2017), 3, https://www.aclu.org/files/
FilesPDFs/ALPR/new-york/alprpra_rensselaercountysheriffdepartment_troyny_2.pdf.
36
addition, the NYPD has established a policy for activating and responding to AMBER
alerts, including specific instructions for manual entries, if required.

d. Other Benefits

Local governments also use LPR technology to track commercial carriers, enable
payment for tolls on roads, bridges, and tunnels to identify locations with heavy traffic, and
estimate travel times to specific highway interchanges. 123 Most toll bridges and tunnels,
in addition to E-Z Pass, now use license plate readers to collect tolls, thus eliminating toll
plazas. Therefore, without lines of stopped vehicles waiting in cash only-lanes, traffic
congestion has eased at area crossings in New York City.

Currently, no laws prohibit private organizations from using LPR technology and
creating their own unique databases for their operations. Commercial uses for LPR
technology include security and management for school campuses, residences, parking,
casinos, airports, and healthcare facilities. Real-time information can identify vehicles that
park in unauthorized or restricted areas as well as vehicles that remain on site longer than
permitted. Many LPR systems connect to security offices in commercial locations that
immediately notify law enforcement if criminal or suspicious activity requires further
investigation.

C. CHALLENGES

Law enforcement agencies encounter challenges in using LPR technology. Both

citizens and activist groups have raised concerns over privacy, data collection, retention,
citizen acceptance, and understanding of the technology, not to mention inaccuracy and
misuse. The New York Civil Liberties Union contends that the NYPD’s five-year data
retention policy for all scanned plates, including plates that are not involved in
investigations, is excessive. 124 In addition, the ACLU argues that the NYPD’s DAS will
track individuals in ways that few citizens can fathom, as the system contains millions of

123 Bryce Clayton Newell, “Local Law Enforcement Jumps on the Big Data Bandwagon: Automated
License Plate Recognition Systems, Information Privacy, and Access to Government Information,” Maine
Law Review 66 (2013): 8.
124 Hirose, “NYPD’s Vast License Plate Reader Database.”

37
data points. Privacy advocates contend that law enforcement agencies can track individuals
using LPR scans from moving vehicles, thereby diminishing levels of privacy. 125
Currently, 14 states, not including New York, have laws that regulate the use of LPR
technology. 126 The NYPD, however, must continue to address the technology’s challenges
to retain public support.

1. Privacy

A primary task in law enforcement is balancing the citizens’ desire for privacy with
the need to keep them safe. Privacy concerns transcend citizen objections over
compromised anonymity from police surveillance technologies. Privacy advocates, such
as the ACLU, express concern over how law enforcement uses surveillance technologies
and how collected data are used, disseminated, and shared with other organizations.
Furthermore, LPRs often raise concerns about misuse, in the forms of implicit bias or weak
policies and procedures.

a. Data Collection/Invasiveness

The main concerns that the ACLU would like to see addressed include law
enforcement’s data collection and retention methods. Kaelyn Rich of the New York Civil
Liberties Union admits, “LPR can be a reasonably useful technology.” 127 Rich also states,
“The problem is they’re storing records that are not hits. They’re keeping these millions of
other records on everyday people going about their business. In the United States, it’s a
core principle that the government does not invade people’s privacy and they do not collect
information on people in case they do something wrong.” 128 New York State has enacted
no laws pertaining to LPRs. Bill S23, however, was proposed for LPRs in 2015 and
modified in 2017 to include permissible use, collection, and retention of information, as

125 Gierlack et al., License Plate Readers for Law Enforcement, 2.

126 Newell, “Local Law Enforcement Jumps on the Big Data Bandwagon,” 402.
127 Steve Orr, “New York Knows Where Your License Plate Goes,” USA Today, July 28, 2014,
https://www.usatoday.com/story/news/nation-now/2014/07/28/new-york-archiving-license-surveillance-
data/13261679/.
128 Orr.

38
well as mandatory open-source policies and annual reporting requirements. 129 The bill
continues to sit in committee with no definitive path for approval. The only existing New
York State document relating to LPRs, the Suggested Guidelines Manual, was produced in
2011. Currently, no universal policy exists in New York State. The NYPD, however, has
created a detailed policy on LPR use, collection, data retention, and training, which the
department has posted on its website.

Merola and Lum have examined the public’s perception of LPR technology,
surveying a diverse group of citizens to categorize the type of LPR use by the degree of
citizen support. 130 The survey revealed that citizens had greater concerns with the
dissemination of collected data, retention time, and storage for wide-ranging use but
showed little to no concern over the use of LPR technology by law enforcement. Overall,
the survey demonstrated support for LPR use—with an existing correlation between
individuals who trust law enforcement and those who support LPR technology.

b. Misuse/Abuse of LPR Technology

Critics raise concerns that officers will misuse LPR technology by, for instance,
using LPR data for personal use, selectively enforcing laws, and sharing data
inappropriately. Critics contend that law enforcement officers may misuse LPR data; the
ACLU argues that officers have the ability to patrol with LPR-equipped vehicles and target
specific individuals, such as persons engaged in prayer inside a mosque. 131 Such tactics,
if conducted at demonstrations, might affect citizens deciding to protest against the
government and, thus, negate their right to free speech. Furthermore, critics contend that
LPR technology can reveal intimate details for which individuals would plead for
anonymity. For example, individuals may be tracked in their vehicles traveling to legal
destinations but engaging in secretive activities. Other privacy concerns include the law
enforcement use of LPR data to collect information on individuals who may systematically

129 “Senate Bill S23,” New York State Senate, last modified December 28, 2016, https://www.
nysenate.gov/legislation/bills/2017/S23.
130 Linda M. Merola and Cynthia Lum, “Understanding Citizen Support for License Plate Readers,”
Translational Criminology (2015): 23, https://www.bja.gov/bwc/pdfs/TC8-Spring2015.pdf#page=25.
131 Crump, You Are Being Tracked, 11.

39
engage in lawful but private or discreet behavior. In addition, officer misuse of the
technology, no matter how infrequent, concerns law enforcement and citizens alike. The
ACLU reports several incidents of officers involving criminal misuse of LPR
technology. 132

c. Ineffective Policies

Many law enforcement agencies have written policies that explain LPR rules for
officers to follow. Policies vary by agency, with different policies for use, collection,
sharing, purging of data, data analysis, and official inquiries into vehicle registrant
information. For example, in 2017, 20 states introduced legislation for LPRs; however, no
laws were passed. 133 In New York State, several of its counties—Nassau, Suffolk,
Westchester, Broome, Albany, Onondaga, Erie, Monroe, and New York City—retain LPR
data for five years.

Also, the ACLU contends that the overall return on investment—LPR reads versus
LPR hits—is less than 1 percent. 134 The ACLU examined 2009–2011 LPR data from the
Minnesota State Patrol and revealed that officers made only 131 arrests and issued 852
citations based on 1.6 million LPR scans. 135 The ACLU contends that most departments
maintain LPR data for long periods, and some do not have policies for LPR data
whatsoever, even if no open investigations or extenuating circumstances exist. 136 With
technology allowing for increased storage capabilities, it is more cost-effective for law
enforcement agencies to retain LPR data for longer periods. Further research will dictate
whether extended retention periods will pay dividends—if more crimes, such as cold cases,
can be solved.

132 Hirose, “NYPD’s Vast License Plate Reader Database.”

133 “Automated License Plate Readers: State Legislation 2016 & 2017,” National Conference of State
Legislatures, accessed May 6, 2018, http://www.ncsl.org/research/telecommunications-and-information-
technology/automated-license-plate-readers-state-legislation-2016.aspx.
134 Crump, You Are Being Tracked, 15.
135 Crump, 15.
136 Crump, 16.

40
 d. Data Storage, Data Sharing, and Hacking

The security of sensitive data challenges government agencies, law enforcement,

and the private sector. LPR information is considered big data, which are defined as “data
sets that are so large or complex that primitive data processing application software is
inadequate to deal with them.” 137 The sheer number of vehicles being scanned on
roadways throughout the United States means LPRs require an abundance of storage space.
In addition, sharing of data across government and law enforcement agencies increases the
potential for cyber-attacks. Recent history demonstrates that cyber-attacks result in serious
damage to company credibility and individual privacy of sensitive, personal information.
Big data breaches—Yahoo in 2013 affecting 1.5 billion user accounts, eBay in 2014
affecting 145 million users, and J.P. Morgan Chase in 2014 affecting 76 million households
and seven million small business—all create trust issues and create massive financial
burdens for both corporations and individuals. 138 Compromised LPR data can reveal
vehicle locations, such as for individuals who frequent locations they wish to keep private,
as well as historical data, such as red light and speed camera violations. For example, in
2015, over 100 LPR cameras from various law enforcement agencies in different states
were left unsecured online, and anyone with an internet connection and a browser had
unlimited access to their data. 139

2. Accuracy

LPR technology faces problems with overall reliability. Examples include dirty
license plates, intentional alteration of plates, and identical vanity plates issued in different
states. 140 Reliability, accuracy, and overall performance concerns are not unique to the use

137 Surbhi Jain, “Big Data: What, Why and Why Not,” International Journal of Engineering
Development 5, no. 2 (2017): 5, https://www.ijedr.org/papers/IJEDR1702334.pdf.
138 Jain, 5.
139 Cooper Quintin and Dave Maass, “License Plate Readers Exposed! How Public Safety Agencies
Responded to Major Vulnerabilities in Vehicle Surveillance Tech,” Electronic Frontier Foundation,
October 28, 2015, https://www.eff.org/deeplinks/2015/10/license-plate-readers-exposed-how-public-safety-
agencies-responded-massive.
140 Gierlack et al., License Plate Readers for Law Enforcement, 15.

41
of LPR technology. Other law enforcement technologies, such as BWCs, experience
similar concerns and are continuously addressed.

LPRs sometimes fail to read damaged or obscured license plates or those covered
with transparent materials, such as clear license plate covers. 141 Mobile LPR readers can
mistakenly extract street addresses and street signs and erroneously record them as license
plates, thereby populating databases with inaccurate data. 142 In the case of extracting LPR
data from large crime scenes, crime analysts spend needless hours sifting through and
disaggregating legitimate information from erroneous data.

a. False Positives, False Negatives, and Misreads

When LPRs scan plates that result in false positives, the overall accuracy of the
collected data can be affected. Although most LPRs are accurate, misreads do occur, and
readers may have difficulty in differentiating between identical license plates from
different states. 143 This may have serious ramifications for both citizens and law
enforcement. For example, if a complainant reports a stolen vehicle with a plate identical
to one issued by another state, it may create a false positive and identify the wrong vehicle
as being stolen. To address this concern in the NYPD, officers are tasked with verifying all
information before taking enforcement action.

b. Intentional Circumvention

Private companies are investing heavily in solutions to circumvent LPR technology

intentionally. Laws that were passed decades ago to prevent intentional circumvention,
such as license plate covers and jammers that scramble the signal emitting from the LPR,
have failed to keep up with new technologies. Some companies claim their devices will
prevent an LPR from scanning the plate. 144 Phantom Plate has designed a spray-on, high-

141 Gierlack et al., 15.

142 Gierlack et al., 15.
143 Gierlack et al., 15.
144 “SunflexZone Anti-ALPR & Red-Light Camera Privacy Solutions,” Sunflex Zone, accessed June
9, 2018, https://www.sunflexzone.com/.
42
powered gloss that over-exposes the images of the license plate and, therefore, intentionally
circumvents the LPR. 145 Other companies sell products, such as clear sprays, that can be
applied over a plate to avoid an LPR. 146 This technology can thwart law enforcement
efforts to enforce speeding, control intersections, and identify stolen vehicles.

3. Transparency and Public Trust

The ACLU recently filed legal action against the U.S. Immigration and Customs
Enforcement (ICE) for failure to provide records on how the agency uses LPR data. 147 The
ACLU expressed concern about how ICE stores and uses the data for civil immigration
enforcement. These concerns have damaging effects on citizen support of law enforcement.
Although ICE has announced the purchase of LPR data on its website and assured citizens
that “the agency complies with privacy and civil liberties requirements,” the organization
explains LPR data are leveraged to “support criminal and administrative law enforcement
missions.” 148 LPR use by ICE is a smaller concern of a larger immigration issue. In
January 2018, 18 citizens, including two New York City council members, were arrested
by the NYPD following a demonstration over the immigration status of an individual.149
The incident highlights the public’s greater focus on how law enforcement collects all data,
including LPR scans, and whether it uses the data against citizens.

145 “Avoid Red Light and Speed Camera Tickets,” Phantom Plate, accessed April 20, 2018,
https://www.phantomplate.com/.
146 “Avoid Red Light and Speed Camera Tickets,” Photo Blocker, accessed April 21, 2018,
https://www.photoblocker.com/photoblocker.html.
147 Sophie Haigney, “ACLU Sues Immigration and Customs Enforcement for License Plate Reader
Records,” SF Gate, May 28, 2018, https://www.sfgate.com/bayarea/article/ACLU-sues-ICE-for-license-
plate-reader-contracts-12937712.php.
148 U.S. Immigration and Customs Enforcement, Acquisition and Use of License Plate Reader Data
from a Commercial Service, DHS-ICE-PIA-039 (Washington, DC: Department of Homeland Security,
December 2017), https://www.dhs.gov/publication/dhs-ice-pia-039-acquisition-and-use-license-plate-
reader-data-commercial-service.
149 Liz Robbins, “Activists and ICE Face Off over Detained Immigrant Leader,” New York Times,
January 12, 2018, https://www.nytimes.com/2018/01/12/nyregion/immigration-activist-deportation.html.
43
 a. Transparency by Law Enforcement

Critics have requested that the NYPD become more transparent with its policing
strategies and policies. In an effort to mitigate those concerns, the NYPD uses social media
applications such as Facebook, Instagram, Twitter, and the department’s webpage to
promote transparency. In addition, the department publishes reports, statistics, and policies
online for public consumption. In March 2018, the NYPD updated its LPR policy and
published it online to reflect the latest advancements of the technology. The policy
explains,

The LPR system allows for the proactive entry of license plate numbers and
partial plate numbers, enabling the system to activate when the wanted
vehicle’s license plate has been read by the LPR device. The Real Time
Crime Center can be contacted to conduct a search of past records of license
plate numbers searched. LPR devices are intended to provide access to
stolen and wanted files and may also be used in furtherance of a criminal
investigation. The use of an LPR device for any other purpose is strictly
prohibited. 150

The new policy informs officers and citizens of its zero-tolerance policy for officer misuse
of LPR technology. While the NYPD has specific guidelines for citizens to obtain BWC
videos to meet citizen expectations of transparency, there are no specific instructions for
citizens to obtain LPR data concerning their vehicles.

In 2014, the NYPD denied a Freedom of Information Law inquiry from the New
York Civil Liberties Union, which requested detailed information on LPR use, collection,
sharing, locations and number of units, as well as the number of scans, officer training, and
EZ-pass data among other concerns. 151 The NYPD denied the request, explaining that the
release of LPR data would interfere with law enforcement investigations and disclose
sensitive techniques and procedures that might endanger the safety of citizens. 152 In
addition, the denial cited Public Officers Law (POL) §87(2)(e)–(g) and (i), which exempt

150 NYPD, Patrol Guide: “Padlock Law.”

151 City of New York, Police Department Legal Bureau, “Freedom of Information Law Request: #14-
PL-0175” (letter to Nate Vogel, New York Civil Liberties Union, April 17, 2014), https://www.nyclu.
org/sites/default/files/20140417_NYPDDenial_NYCLUFOILRequest.pdf.
152 City of New York, Police Department Legal Bureau.

44
such records from release. 153 Prioritizing safety and remaining transparent is challenging
for law enforcement, especially when the public interprets surveillance technologies as
deceptive tactics.

b. Community Mistrust and Acceptance

There seems to be a greater degree of acceptance for LPRs than other law
enforcement surveillance technologies. The NYPD created and updated strong policies and
procedures governing the use of LPR technology, which mitigated most concerns of
mistrust among the community. Growing public acceptance and increased transparency
have contributed to the capabilities of LPR technology because citizens today are better
informed and more apt to challenge traditional methods of policing. Citizens have been
aware of the benefits of LPR technology, such as its use in expediting traffic flow and
easing congestion on area bridges and tunnels, ever since toll collection booths were
replaced with fixed LPRs. Nevertheless, LPR systems work in conjunction with other
technologies that are not governed by police departments, creating the potential for new
privacy concerns.

In June 2018, California began piloting license plates that digitally display the plate
number as well as commercial advertisements; the digital plates also transmit location data
for future analysis and record driving habits. 154 The benefits are that the plate pays tolls
automatically, tracks stolen vehicles, tracks mileage, and receives virtual boundary
notifications. This new technology also receives notification from such public service
messaging as street closures and AMBER alerts. These benefits, however, may erode
personal privacy. Stephanie Lacambra from the Electronic Frontier Foundation contends
that vehicle location data have “the potential to reveal a lot more than . . . where you happen
to be at a particular moment in time.” 155 Although the capabilities of the digital license

153 City of New York, Police Department Legal Bureau.

154 James Doubek, “Digital License Plates Roll Out in California,” NPR, June 1, 2018,
https://www.npr.org/sections/thetwo-way/2018/06/01/616043976/digital-license-plates-roll-out-in-
california.
155 Doubek.

45
plate may sacrifice citizen privacy by sharing all of the metadata associated with a vehicle,
the pilot is voluntary and offers benefits for both citizens and law enforcement.

D. CONCLUSION

LPR technology continues to evolve, thereby creating additional uses for law
enforcement and the private sector. Initially, LPR technology enabled law enforcement to
track stolen vehicles. New uses of LPR technology include toll collection, summons
enforcement, and monitoring of traffic conditions. LPR technology assists law enforcement
in reducing and investigating crimes and terror attacks, increasing officer safety, locating
missing persons, and supporting individuals with sensitive needs. In addition, private
companies are using LPR technology to inventory vehicles parked in lots and garages, as
well as scanning plates to identify scofflaws on the road or vehicles subject to
repossession. 156 Roberts and Casanova contend that LPR technology has a positive impact
on the law enforcement community and deems it successful in multiple uses. 157

While no technology exists without some shortcomings, LPR has proven beneficial
to law enforcement despite some systematic flaws. An amalgamation of FRT and LPR
technologies may appear in the near future, as suggested with the high-occupancy vehicle
example. In the near the future, FRT will likely be used to identify faces in high-occupancy
vehicle lanes on area highways to determine whether the appropriate number of people are
inside the vehicle. 158 It is unknown, however, how law enforcement will construct
additional methods to leverage both LPR technology and real-time FRT for the benefit of
the public and law enforcement.

In exchange for improved traffic conditions, though, citizens may begin to feel
diminished levels of privacy—yet such LPR applications as summons enforcement and

156 Kaveh Waddell, “How License-Plate Readers Have Helped Police and Lenders Target the Poor,”
Atlantic, April 22, 2016, https://www.theatlantic.com/technology/archive/2016/04/how-license-plate-
readers-have-helped-police-and-lenders-target-the-poor/479436/.
157 Roberts and Casanova, Automated License Plate Recognition.
158 Dave Maass, “The Four Flavors of Automated License Plate Reader Technology,” Electronic
Frontier Foundation, April 6, 2017, https://www.eff.org/deeplinks/2017/04/four-flavors-automated-license-
plate-reader-technology.
46
crime fighting mutually benefit law enforcement and citizens. Furthermore, citizens are
living in a society connected by technology, social media, and the internet. The paradigm
that defines their acceptance of new privacy levels has yet to be determined.

47
THIS PAGE INTENTIONALLY LEFT BLANK

48
 III. FACIAL RECOGNITION TECHNOLOGY

Like LPR technology, FRT—especially when used in real time—has the advantage
of acting as a force multiplier for the NYPD in crowded environments such as the New
York City subway system. LPRs and FRT are similar technologies, so they share many of
the same advantages and disadvantages. The NYPD mitigated many of the challenges of
LPRs and could ostensibly apply the same framework to address the challenges of real-
time FRT. This chapter provides an overview of FRT, including the history and definition
of the technology, reviews FRT applications, and describes how the benefits will outweigh
the risks. Finally, it addresses the concerns over how law enforcement will collect, share,
and disseminate personal information obtained from facial recognition software.

A. OVERVIEW

In the 1960s, Woodrow W. Bledsoe developed the first semi-automatic face

recognition system. 159 This method identified the ears, eyes, nose, and mouth on
photographs and then calculated the distances and ratios to compare features to source
reference data. 160 FRT processes and matches the unique characteristics for identification
or authorization. FRT uses a digital or video camera that enables the software to detect
images of unknown individuals, analyze the features within the images, and compare those
analyzed features to images of known individuals within a database. 161 FRT is similar to
other biometrics, such as fingerprints, iris scans, and voice recognition.

Fifty years after its invention, FRT has integrated into our everyday lives, creating
new societal benefits around the world. With recent advancements in FRT, law
enforcement, the federal government, and the private sector all look to leverage the
technology to make their operations more secure, more efficient, and—in some cases—

159 Woodrow Wilson Bledsoe, Proposal for a Study to Determine the Feasibility of a Simplified
Facial Recognition Machine (Palo Alto, CA: Panoramic Research, January 30, 1963), http://archive.org/
details/
firstfacialrecognitionresearch.
160 Bledsoe.
161 Find Biometrics, “Facial Recognition.”

49
more profitable. The greatest change in FRT in the last decade is its ability to accurately
identify persons in real time. While real-time technology has the potential to change
modern policing with significant public benefits, leadership must proactively address the
resulting challenges.

B. BENEFITS OF FRT

Facial recognition is a non-intrusive biometric technology, meaning that facial

recognition systems can scan faces without citizen participation and, when used in real
time, identify people as they walk. Citizens do not need to stop walking or look directly
into a camera, so real-time FRT can act as a force multiplier for the NYPD in the New
York City subway system by monitoring the ridership in all 472 stations. If real-time FRT
existed in the subway, the NYPD could instantly screen millions of passengers each day—
identifying both wanted criminals and suspected terrorists—but without unnecessary
personal intrusions, such as bag checks and container screenings, during peak operational
hours. Public safety benefits accentuate the greater good that real-time FRT can provide
law enforcement. Figure 2 depicts how FRT algorithms match images against a database.

50
 Figure 2. Algorithm for Matching Facial Images 162

In law enforcement, leadership needs to evaluate technology when opportunities

arise to make officers more efficient, better informed, and safer. Allevate, a company that
promotes enhanced safety and operational efficiency, contends that recent acts of terrorism
around the world demonstrate less sophistication but increased frequency. 163 The company
argues that technology can play a critical role in improving the efficiency of law
enforcement and intelligence agencies and furthering the ability to interrupt future terror
plots. 164 Technology has changed the way people think of traditional business models.
Throughout history, law enforcement has often been astute in using technology to combat
crime. For example, law enforcement has leveraged technologies, such as the automobile,
the telephone, and more recently, biometric technology to create significant efficiencies.
According to one law enforcement scholar, technology in policing “helped emphasize

162 Source: Jennifer Lynch, Face Off, Law Enforcement Use of Face Recognition Technology (San
Francisco: Electronic Frontier Foundation, February 12, 2018), 5, https://www.eff.org/wp/law-
enforcement-use-face-recognition.
163 “Helping to Counter the Terrorist Threat Using Face Recognition,” Allevate, July 28, 2017,
http://allevate.com/index.php/2017/07/28/helping_to_counter_the_terrorist_threat_using_face_recognition/.
164 Allevate.

51
discipline, equal enforcement of the law and centralized decision making.” 165 Until 10
years ago, it would have been crazy to think of starting a hotel business without hotels, a
taxi company without cabs, or a merchandise company without stores in which to shop,
but today such innovators as Airbnb, Uber, and Amazon have changed the way companies
think and function. 166 These same principles apply to law enforcement, which requires
abstract thinking to solve problems, create additional efficiencies, and foster stronger
community relationships. Figure 3 indicates the process by which FRT generally operates.

165 Seaskate, The Evolution and Development of Police Technology (Washington, DC: Seaskate, July
1, 1998), https://www.ncjrs.gov/pdffiles1/Digitization/173179NCJRS.pdf.
166 William D. Eggers, Delivering on Digital: The Innovators and Technologies That Are
Transforming Government, 1st ed. (New York: Rosetta Books, 2016), loc. 79–121.
52
 Figure 3. How FRT Systems Generally Work 167

1. Crime Prevention Benefits

The most important benefits derived from the use of FRT all involve preventing or
reducing crime. Although the use of FRT already benefits the NYPD, if the department

167 Source: Government Accountability Office, Facial Recognition Technology: Commercial Uses,
Privacy Issues, and Applicable Federal Law, GAO-15-621 (Washington, DC: GAO, July 2015).
53
uses FRT in real time, crime reduction benefits—potentially terrorism prevention and rapid
wanted person identifications—grow exponentially.

a. Preventing Violent Crime

Law enforcement began piloting real-time FRT for large events and, over time,
expanded its use at demonstrations and transportation systems. The first use of real-time
FRT in the United States occurred during the Super Bowl in 2001; law enforcement
screened 100,000 people using FRT—among whom 19 people were arrested due to active
arrest warrants. 168 This pilot was the first of its kind at a high-profile event. More recently,
police are beginning to revisit the idea of using real-time FRT at large, high-profile events.
In 2017, the Metropolitan Police in London deployed real-time FRT to scan crowds at a
remembrance ceremony. The database contained 50 individuals known for exhibiting
compulsive behavior toward public figures. Although none of the individuals was wanted
for prior crimes, it was a proactive approach to mitigating potential safety concerns at large-
scale, high-profile events. 169

In more modern times, real-time FRT has the capability to stop criminals who
commit serial violent crimes. On May 13, 2015, David Baril attacked NYPD Police Officer
Lauren O’Rourke with a hammer in midtown Manhattan when O’Rourke and her partner
approached Baril to arrest him. He was wanted for a string of serious robberies in
Manhattan, striking his victims in the head with a hammer and attempting to remove
property. 170 Baril had already been a wanted criminal with an active arrest warrant prior
to his recent pattern of robberies. Following his latest robbery, NYPD detectives combed
the area for clues. After examining CCTV videos of the Herald Square Subway Station’s
platforms and turnstiles, detectives observed an individual on video inside of the subway

168 Daniele Cavadini et al., “Introducing the Biometrical Electronic Passport (EPass)” (seminar,
University of Fribourg, 2008), 8.
169 Mark Townsend, “Police to Use Facial-Recognition Cameras at Cenotaph Service,” Guardian,
November 11, 2017, http://www.theguardian.com/technology/2017/nov/12/metropolitan-police-to-use-
facial-recognition-technology-remembrance-sunday-cenotaph.
170 J. David Goodman and Al Baker, “Police Shoot Hammer-Wielding Man Sought in 4 Manhattan
Attacks,” New York Times, May 13, 2015, https://www.nytimes.com/2015/05/14/nyregion/officer-shoots-
man-in-midtown-manhattan.html.
54
station that matched a description provided by the victim. The robbery victim was able to
identify the photo as the individual who attacked him.

The NYPD’s Facial Identification Section identified Baril using FRT and
disseminated his picture onto the department smartphones issued to every police officer.
Two days later, Baril was walking in midtown Manhattan when Officer O’Rourke observed
him. While FRT was attributed with Baril’s initial identification, real-time FRT could have
identified Baril earlier as a wanted criminal, which would have resulted in his immediate
apprehension and thereby protected citizens from additional acts of violence. One of the
primary benefits of real-time FRT in contrast with FRT, as a post-investigative crime-
fighting strategy, is the ability for law enforcement to identify wanted career felons and
apprehend them in crowded environments before they commit additional violent crimes.

b. Investigating and Preventing Acts of Terrorism

Real-time FRT can also identify and alert police officers to the presence of a known
terrorist entering or exiting a specific location. In the New York City subway system, a
real-time system would enable NYPD police officers to screen millions of people each day
and potentially identify individuals on terror watch lists. Other parts of the world are
evaluating real-time FRT systems. In Germany, police in Berlin are testing real-time FRT
in the Berlin Train Station, contending that the technology can alert police of known
terrorism suspects. 171

Although no acts of terrorism have been successfully prevented using real-time

FRT, law enforcement professionals sense that because of recent technological
advancements, using real-time FRT may prevent an act of terrorism. For example,
following the London terror attacks at the Borough Market on June 28, 2017, former MI5
Chief Johnathan Evans claimed, “The attacks . . . were very likely preceded by
reconnaissance activity, and one can envisage that the use of FRT would be able to identify

171 Justin Huggler, “Facial Recognition Software to Catch Terrorists Being Tested at Berlin Station,”
Telegraph, August 2, 2017, http://www.telegraph.co.uk/news/2017/08/02/facial-recognition-software-
catch-terrorists-tested-berlin-station/.
55
that in advance and enable preemptive action to be taken.” 172 Although several law
enforcement agencies around the world are currently testing real-time FRT, further
research is necessary before both law enforcement and facial recognition software
companies can assert conclusive findings of its effectiveness.

c. Real-Time Identifications of Wanted Persons and Known Suspects

FRT is increasingly being used as a real-time application to identify wanted persons

and known suspects. Instant identifications of individuals have the potential benefits of
making the airline industry safer, locating missing children and adults, and helping
instantly to apprehend individuals wanted for serious crimes. Several companies contend
that real-time FRT is advanced enough for use on a variety of platforms, for the military,
law enforcement, and the private sector. Technology companies NEC (NeoFace Watch)
and Cognitec both specialize in real-time FRT and contend that their facial recognition
software is accurate with real-time functionality, regardless of various angles and
lighting. 173 According to NEC, NeoFace Watch mitigates challenges such as “crowded
environments, poor lighting, moving subjects and multiple variables as small yet
significant as spectacles, hats, and scarves.” 174

An additional benefit of the real-time use of FRT is the ability to identify wanted
and missing persons. In November 2016, Safran Identity & Security, formerly Morpho,
collaborated with INTERPOL to create a database of facial images. Both parties recognized
the opportunities that this ever-expanding biometric science brought to law enforcement.
Now there is a tool available to law enforcement that can be used globally to identify
missing and wanted persons, as well as persons of interest. INTERPOL intends to take the

172 “Facial Recognition Technology ‘Could Prevent Future Terror Attacks,’” ITV News, March 1,
2018, http://www.itv.com/news/london/2018-03-01/facial-recognition-technology-could-prevent-future-
terror-attacks/.
173 Rawlson O’Neil King, Special Report: Biometrics in Law Enforcement (Lansing, MI: Biometrics
Research Group, Michigan State University, 2017), 1–3, https://www.biometricupdate.com/wp-content/
uploads/2017/08/special-report-biometrics-in-law-enforcement.pdf.
174 NEC Global Face Recognition Centre of Excellence, NeoFace Watch, High Performance Face
Recognition (NEC Global Face Recognition Centre of Excellence, 2016), https://www.nec.com/en/global/
solutions/safety/face_recognition/PDF/Face_Recognition_NeoFace_Watch_Brochure.pdf.
56
project to the next phase by enhancing in-field operations, allowing certain photos to be
accessible on smart devices. Once it enacts this change, INTERPOL can use that data to
check faces against specific watch lists in real time. 175

In addition, major transportation facilities in the United States are beginning to test
FRT. On October 11, 2017, U.S. Customs and Border Protection began a pilot program
using FRT to screen passengers arriving and departing John F. Kennedy Airport in New
York City. 176 John Wagner, deputy executive assistant commissioner of the Office of Field
Operations, explains, “As we continue to deploy technical demonstrations, CBP is
assessing the use of biometric technology as part of a future end-to-end process, from
check-in to departure, in which travels use biometrics instead of their boarding pass or ID
throughout the security and boarding process.” 177

d. Investigative Assistance

The NYPD has the capability to leverage FRT for use in real time, but currently,
the NYPD is not using any form of real-time FRT. The department does use FRT, however,
as a post-investigative tool to identify perpetrators involved in crimes and to identify
deceased and unidentified persons. The NYPD’s Facial Identification Section (FIS) has
existed since 2011 and operated as a subunit under the Real Time Crime Center. It operates
as an investigative mechanism for detectives in the field, and their mission is to provide
investigative leads to assist detectives in identifying unknown individuals related to
investigations through biometric algorithms, intelligence data, social media, and any other
investigative means. Individuals with moderate disfigurements can still be analyzed and
identified by detectives using FRT software. Precinct commanders and detective squad
commanders utilize the resources of FIS to aid in criminal investigations in which the
perpetrator is unidentified. In addition, FIS detectives leverage social media to help identify

175 “Forensics,” INTERPOL, accessed July 3, 2018, https://www.interpol.int/INTERPOL-expertise/

Forensics/Facial-recognition.
176 “CBP Deploys Facial Recognition Biometric Technology at 1 TSA Checkpoint at JFK Airport,”
Customs and Border Protection, October 11, 2017, https://www.cbp.gov/newsroom/national-media-release/
cbp-deploys-facial-recognition-biometric-technology-1-tsa-checkpoint.
177 Customs and Border Protection.

57
unknown perpetrators. Facebook and Instagram have a database of pictures larger than that
of all Department of Motor Vehicles (DMV) and FBI databases combined. 178

FIS works with the private sector and other city agencies to educate them on camera
placement, video quality, and methods used to read faces, so FRT software can successfully
process and analyze the image. FIS, however, is still subject to scrutiny. In 2018,
Georgetown University’s Center on Privacy and Technology filed a lawsuit against the
NYPD, stating the department’s “face-recognition system appears to include data for every
NYPD arrestee, meaning that each arrestee is subjected to face-recognition searches.” 179
In addition to the NYPD database, the New York State DMV maintains a massive photo
database. Although several states use their respective DMV databases to conduct searches,
the NYPD does not currently use DMV photographs for database comparisons. The New
York State DMV, however, as well as the Department of Motor Vehicles in 39 other states
use FRT to categorize over 16 million photos to assist in detecting fraudulent licenses. 180
In New York State, the program began in January 2017, and Governor Cuomo says it “has
led to the arrest of 100 suspected identity thieves and opened an additional 900 unsolved
cases.” 181 There is strong evidence to show that as FRT evolves, criminals are easier to
detect and the evidence required in establishing probable cause is easier to obtain.

178 Tom Risen, “Could the FBI See Your Selfies?,” U.S. News & World Report, July 8, 2014,
https://www.usnews.com/news/articles/2014/07/08/fbi-may-seek-facebook-data-for-facial-recognition.
179 Steven Rex Brown, “NYPD Ripped for Abusing Facial-Recognition Tool,” New York Daily News,
March 1, 2018, http://www.nydailynews.com/new-york/nyc-crime/nypd-ripped-abusing-facial-recognition-
tool-article-1.3847796.
180 David Kravets, “Enhanced DMV Facial Recognition Technology Helps NY Nab 100 ID Thieves,”
Ars Technica, August 28, 2016, https://arstechnica.com/tech-policy/2016/08/enhanced-dmv-facial-
recognition-technology-helps-ny-nab-100-id-thieves/.
181 “Governor Cuomo Announces More Than 100 Arrests since Major Enhancement to DMV’s Facial
Recognition Technology,” New York State Governor’s Office, August 24, 2016, https://www.governor.
ny.gov/news/governor-cuomo-announces-more-100-arrests-major-enhancement-dmvs-facial-recognition-
technology.
58
 2. Law Enforcement and Public Benefits

In addition to crime-prevention benefits, real-time FRT may increase officer safety,

assist officers in the field when interacting with individuals who cannot care for
themselves, and help locate missing persons.

a. Officer Safety

Although no specific research correlates FRT to officer safety, the possibility exists
that if FRT can integrate with other technologies, such as LPR and BWC technologies,
officers will have more information when interacting with individuals, such as persons
wanted for murder or who have prior arrests for assaulting police officers. In addition,
officer stops can be expedited if citizen identifications are validated in the field. For
example, Transit NYPD officers sometimes wait 30–45 minutes for a marked patrol vehicle
to transport an individual back to the station house solely to ensure proper identification
prior to issuing a summons. If FRT could properly identify a person in the field, a police
officer can issue a summons on the scene, thereby saving valuable time for both citizens
and police officers. In addition, officers face increased risks when waiting for a vehicle to
transport the perpetrator. Increased wait times may lead to avoidable confrontations,
officers being assaulted, or perpetrators attempting to escape custody.

b. Mental Health Support

In addition to crime-fighting benefits, FRT—especially when used in real time—

can assist officers in helping individuals who might be unable to care for themselves. FRT
can immediately alert officers of potential special needs of the individual with whom they
are interacting. In California, the San Diego County Sheriff’s Department uses FRT to
assist citizens with disabilities such as autism, dementia, Alzheimer’s, down syndrome, or
any other condition. 182 The Take Me Home Program provides officers with emergency
contact information, physical conditions, and special needs to help the individual. 183 FRT

182 “Home Page,” San Diego County Sheriff’s Department, accessed May 29, 2018, 1–2, http://www.
sdsheriff.net/tmh/docs/tmh-english.pdf.
183 San Diego County Sheriff’s Department, 2.

59
provides officers with potentially life-saving information in the event that the individual is
unable to communicate. For example, if officers encounter an unconscious person, FRT
could search the database for a match and ascertain critical information, including but not
limited to allergies, use of a pacemaker, or contact information for a relative.

c. Missing Persons

Law enforcement could benefit from the use of real-time FRT if missing persons
were included in the database. In 2017, there were 651,226 persons reported missing in the
United States. 184 In 2009, for example, 13-year-old Francisco Hernandez Jr. was reported
missing in New York City. Hernandez, diagnosed with Asperger syndrome, spent 11 days
in the New York City subway system before he was finally located by a police officer. 185
Hernandez was underground most of the time, wandering around subway stations and
riding different trains. He was ultimately identified by a transit police officer, who
recognized the boy from missing person posters in the subway. There is a strong possibility
that, if real-time FRT had existed in the New York City subway system, Hernandez could
have been located sooner. If the NYPD had used real-time FRT, the NYPD detectives could
have entered the boy’s photograph into the facial recognition database from the moment
the boy’s mother reported her son missing. If Hernandez had walked by any of the
thousands of cameras located in the subway system, real-time FRT would have alerted
police to his exact location.

In 2017, a six-year-old girl was reported missing in China, and the only
documentation provided to police was an old photo, taken several years earlier. Police were
able to locate the missing girl quickly through an advanced web of cameras linked with
real-time FRT. 186 The system, produced by Skynet, is 99.8 percent accurate and can scan

184 “2017 NCIC Missing Person and Unidentified Person Statistics,” Federal Bureau of Investigation,
accessed June 17, 2018, https://www.fbi.gov/file-repository/2017-ncic-missing-person-and-unidentified-
person-statistics.pdf/view.
185 Jamie Guzzardo and Jesse Solomon, “Missing Boy Spent 11 Days Wandering New York
Subways,” CNN, November 25, 2009, http://www.cnn.com/2009/US/11/25/new.york.subway.teen/
index.html.
186 “‘Skynet’ System Supported by Facial Recognition Technology Boosts Chinese Public Safety”
People’s Daily Online, March 26, 2018, http://en.people.cn/n3/2018/0326/c90000-9441798.html.
60
faces regardless of angles and lighting. In addition, Skynet claims exceptional system
performance in apprehending perpetrators who commit violent crimes such as homicides,
robberies, or abductions. 187

d. Overcrowding in Transit

One of the many challenges for subway riders entering the New York City subway
system is its antiquated payment systems and burdensome points of entry. Citizens often
experience long lines, especially during morning and afternoon rush hour and when
purchasing a daily, weekly, or monthly fare, known as a MetroCard. Individuals can
purchase MetroCards from vending machines or from booth clerks, located near subway
entrances. The current system creates long lines for frustrated customers waiting to
purchase a MetroCard, who are then forced to wait while each person swipes for access to
the subway. In addition, long lines can create opportunities for criminals. Pickpocketing
and other similar crimes occur when customers juggle multiple items, such as wallets, cell
phones, credit cards, or cash, in order to purchase a MetroCard from a vending machine.

Just as technology has advanced from the subway token to the MetroCard, the next
revolutionary payment method could become real-time FRT. A company in the United
States began testing real-time FRT for instant payment and access into subway systems.
Such a system would eliminate the need for MetroCard vending machines and cumbersome
turnstiles and gates. 188 Long lines caused by individuals with expired or damaged
MetroCards, or those unfamiliar with their use, would be replaced with an open
walkthrough area that processes payments as persons walk into the subway system. This
technology is similar to recent changes made at New York City tunnels and bridges. Toll
plazas were eliminated at all bridges and tunnels, reducing traffic congestion. Vehicle
payments are processed through E-Z Pass or LPRs, which photograph the license plate.
Figure 4 illustrates the notional use of FRT in mass transit settings.

187 People’s Daily Online.

188 Tammy Waitt, “CUBIC Testing Face Recognition for Subway (Learn More, Video),” American
Security Today, October 10, 2017, https://americansecuritytoday.com/cubic-testing-face-recognition-
subway-learn-video/.
61
 Figure 4. The Gateless Gate Line Concept Using FRT 189

e. Integration with Other Technologies

FRT could also be integrated with other existing law enforcement technologies such
as LPRs, BWCs, drones, and smartphones. For example, New York State legislators are
evaluating the potential for LPRs to work in tandem with FRT in New York State tunnels
and bridges. 190 A vast and interconnected camera system would identify vehicles and
persons entering and exiting New York City. Lynn Gore, deputy chief procurement officer
for the Metropolitan Transit Authority (MTA)’s bridges and tunnels, issued a memo to
request facial detection and recognition at all Triborough Bridge and Tunnel Authority
facilities, stating,

189 Source: Waitt.

190 “Memo: New York Called for Face Recognition Cameras at Bridges, Tunnels,” Vocativ, last
modified January 27, 2017, http://www.vocativ.com/396745/memo-new-york-called-for-face-recognition-
cameras-at-bridges-tunnels/.
62
 The Authority is interested in implementing a Facial Detection System, in
a free-flow highway environment, where vehicle movement is unimpeded
at highway speeds as well as bumper-to-bumper traffic, and license plate
images are taken and matched to occupants of the vehicles (via license plate
number) with Facial Detection and Recognition methods from a gantry
based or road-side monitoring location. 191

The MTA recently constructed towers to house homeland security equipment but would
not reveal the capabilities of the towers—or whether the towers include FRT. 192 The
integration of surveillance technologies, however, clearly illustrate that government
agencies are evaluating various technologies to determine overall efficacy.

Drones can be enabled with FRT to identify persons on the ground for surveillance,
search for specific individuals, and monitor such remote locations as large outdoor
events. 193 These functions would significantly benefit law enforcement because drones
can access densely populated events—in particular, New Year’s Eve celebrations at Times
Square in New York City—to scan crowds for suspicious persons at and near the event.
BWCs outfitted with real-time FRT also have benefits despite the challenges and additional
research needed to understand their long-term effects on society. In addition to personal
safety, officers would have real-time information of persons with arrest warrants or
individuals on FBI terror watch lists.

The NYPD equips each of its approximately 36,000 officers with a smartphone.
Department phones give officers real-time data such as information on wanted persons,
active warrants in multi-dwelling locations, the history of a location, 9–1-1 calls,
ShotSpotter information, and many other functions. In a recent arrest, NYPD officers
recovered two firearms and arrested three individuals after responding to a report of shots

191 Lynn Gore, “Request for Information (RFI), RFI-16-63, All-Electronic Facial Detection and
Recognition System at all TBTA Facilities” (official memorandum, MTA Bridges and Tunnels, December
12, 2016), https://www.documentcloud.org/documents/3428597-RFI.html.
192 Andrew Siff, “New MTA Towers Can Read License Plates, and Maybe More,” NBC (New York),
September 28, 2017, http://www.nbcnewyork.com/news/local/MTA-Bridge-and-Tunnel-Gateway-Towers-
Can-Read-License-Plates-Security--448568193.html.
193 Hwai-Jung Hsu and Kuan-Ta Chen, “Face Recognition on Drones: Issues and Limitations” in
DroNet ‘15: Proceedings of the First Workshop on Micro Aerial Vehicle Networks, Systems, and
Applications for Civilian Use (New York: ACM Press, 2015), 39, http://dl.acm.org/citation.cfm?
doid=2750675.2750679.
63
being fired. Responding officers had used their smartphones to ascertain whether
individuals with active warrants were at the location. 194 If the NYPD uses FRT on
department smartphones, officers could ascertain the identity of individuals without
identification by taking a photograph and using the picture to compare the person to those
contained in a database.

C. CHALLENGES

Law enforcement anticipates several potential challenges in using FRT. The NYPD
must address privacy, accuracy, and public trust of real-time FRT and gain public
acceptance to adopt real-time FRT successfully as a viable law enforcement tool.

The ability of law enforcement to mitigate privacy concerns remains uncertain

partially because no universal policies exist for collecting, sharing, and disseminating facial
recognition data. A New York University study argues that law enforcement agencies
should focus on the ethical issues raised with sharing data because expanding the use of
new technologies compromises the desire of an individual to have complete and absolute
privacy. 195 Although it is important for law enforcement agencies to focus on the moral
and ethical concerns of citizens, it is vital to understand how citizens want to be policed.

Inaccuracy, blurred or partial images, intentional circumvention of facial

recognition cameras, and other factors may affect the identification of wanted persons or
persons of interest. Although FRT is increasingly more accurate than ever before, it is still
imperfect, and as such, several variables will inevitably contribute to false positives, false
negatives, and undetected faces. Technology company NEC has developed algorithms that
utilize the “generalized matching face detection method,” which “provides high speed and
high accuracy for facial detection and facial features extraction.” 196

194 “Finest Technology Helps New York’s Finest Take Two Guns off the Streets,” NYPD News,
December 8, 2015, http://nypdnews.com/2015/12/finest-technology-helps-new-yorks-finest-take-two-gun-
off-the-streets/.
195 Lane et al., Privacy, Big Data, and the Public Good, 49.
196 “Technology,” NEC Corporation, accessed April 15, 2018, http://www.nec.com/en/global/
solutions/safety/Technology/FaceRecognition/index.html.
64
 Negative public perception epitomizes the challenges of real-time FRT and its use
by law enforcement. Public trust of police has deteriorated in recent years and must be
rebuilt to establish communities that work together with the police. Some potential
community concerns over the use of real-time FRT are relevant in this discussion. 197 One
specific concern expressed by citizens is that FRT will increase racial bias toward
minorities when used by law enforcement. Although the FBI states in its report that FRT
is not biased in any way—because it relies on existing data on criminals—the technology
still identifies minorities, African Americans particularly, at a higher rate than other
races. 198 Increased stops without probable cause to arrest may increase situations where
police use force against minorities and worsen the fragile ecosystem of limited trust that
has taken decades to build between law enforcement agencies and minority communities.

1. Privacy

Privacy is the leading concern for citizens involving surveillance technologies used
by law enforcement. Nissenbaum and Introna contend that privacy is one of the “most
prominent concerns raised by critics of FRT.” 199 Real-time FRT can operate clandestinely,
so the potential uncertainty surrounding its use of identifying information is unnerving for
most citizens. Privacy, as defined by Ruth Gavison, is a “measure of the access others have
to you through information, attention, and physical proximity.” 200 Informational privacy,
or control over personal information, is the principal concern involving FRT, due to a
citizen’s lack of control over personal information—and the wish to remain anonymous
regardless of the technology’s security benefits. The security and safety of the people,

197 Trevor Kapp, “Neighborhood Policing Changing Attitudes and Reaping Benefits, NYPD Says,”
DNA Info, April 6, 2017, https://www.dnainfo.com/new-york/20170406/central-harlem/nypd-
neighborhood-policing-nco-community-police-department.
198 Nellie Bowles, “‘I Think My Blackness Is Interfering’: Does Facial Recognition Show Racial
Bias?,” Guardian, April 8, 2016, http://www.theguardian.com/technology/2016/apr/08/facial-recognition-
technology-racial-bias-police.
199 Lucas Introna and Helen Nissenbaum, Facial Recognition Technology: A Survey of Policy and
Implementation Issues (New York: Center for Catastrophe Preparedness and Response, 2010), 44,
https://www.nyu.edu/projects/nissenbaum/papers/facial_recognition_report.pdf.
200 Ruth Gavison, “Privacy and the Limits of Law,” Yale Law Journal 89, no. 3 (1980): 434–435,
https://www.jstor.org/stable/pdf/795891.pdf.
65
however, remain the primary concern of law enforcement agencies, as they are part of
every officer’s sworn duty to serve and protect.

The definition of privacy is often conceptual with various meanings loosely based
on Fourth Amendment rights. Norma Mollers and Jens Halterlein, who examine privacy
issues in public discourse, contend, “In dealing with surveillance, scholars have widely
agreed to refute privacy as an analytical concept. Nonetheless, in public debates,
surveillance technologies are still confronted with issues of privacy, and therefore, endured
as an empirical subject of research on surveillance.” 201 Privacy will always be a topic of
concern, especially because of the new paradigm shift toward technology-based policing.

Real-time FRT in the subway system moving more than six million people per day
will analyze unprecedented amounts of information about citizens. Without specific laws
governing law enforcement’s use of FRT, many contend that databases will identify people
at the whim of law enforcement and, as a result, violate the constitutional rights of
citizens. 202 Some scholars argue that FRT will increase the likelihood of racial profiling
and greater scrutiny among particular races, ethnicities, or religions. A New York
University study suggests, “Historically affected racial groups will generate increased false
positives based upon the methods in how certain systems identify individuals.” 203 Because
of these performance biases, individuals with prior negative interactions with police may
see FRT as only another form of racial discrimination. The propensity for bias in policing
has always been a topical concern, but with biometric technologies rapidly evolving and
being used by law enforcement, state legislators are quickly trying to create new laws to

201 Norma Möllers and Jens Hälterlein, “Privacy Issues in Public Discourse: The Case of ‘Smart’
CCTV in Germany,” Innovation: European Journal of Social Sciences 26 (March 1, 2013): 65,
https://publishup.uni-potsdam.de/opus4-ubp/frontdoor/deliver/index/docId/7767/file/moellers_diss.pdf#
page=73.
202 Kanya A. Bennett, “Can Facial Recognition Technology Be Used to Fight the New Way against
Terrorism: Examining the Constitutionality of Facial Recognition Surveillance Systems,” North Carolina
Journal of Law & Technology 3 (2001): 151, http://ncjolt.org/can-facial-recognition-technology-be-used-
to-fight-the-new-war-against-terrorism-examining-the-constitutionality-of-facial-recognition-surveillance-
systems/.
203 Introna and Nissenbaum, Facial Recognition Technology, 45.

66
build on the Illinois Biometric Act, which focuses on protecting individual privacy. 204 The
goal of the new legislation is an attempt to keep pace with the rapid evolvement of
biometric technology.

Technological advancements with facial recognition have outpaced legislation, and

private organizations are utilizing FRT to perform a variety of functions—corporations
large and small are creating personalized databases and developing new uses for the data,
all with very few restrictions. Martin Abrams, president of the Centre for Information
Policy Leadership at Hunton & Williams, states, “The nature of what we can capture about
individuals is expanding faster than our ability to think about whether it’s prudent to do
so.” 205 Author Muzamil Riffat explains,

Although the notion of privacy as a right does not specifically appear in the
U.S. constitution, it can be deduced from other related provisions. The
amendments made to the Constitution afterward are understood to have
addressed the concerns related to the protection of privacy. These are the
First (speech, religion), Third (quartering soldiers), Fourth (against seizure
and searches), Fifth (against self-incrimination), ninth (for general liberties)
and fourteenth amendments (for personal liberty versus state action). 206

The Fourth Amendment has the strongest role in determining loose guidelines for what is
accepted as reasonable if FRT is used in public spaces. 207 The argument can be made,
however, that public places diminish levels of privacy. Professor Peter Squire explains,

204 Mark Melodia, Paul Bond, and Angela-Angelovska-Wilson, “Legal Risks and Rules of the Move
to Biometrics,” New York Law Journal, March 2, 2015, https://www.technologylawdispatch.com/wp-
content/uploads/sites/26/2016/02/Legal-NYLJ-Article-Risks-and-Rules-of-the-Move-to-Biometrics.pdf.
205 Melissa Maleske, “Facial Recognition Presents Privacy Concerns,” Inside Counsel, March 2012,
https://www.law.com/almID/4f46a3f8150ba0c00a000048/.
206 Muzamil Riffat, “Legal Aspects of Privacy and Security: A Case- Study of Apple versus FBI
Arguments,” Sans Institute, June 1, 2016, 5, https://www.semanticscholar.org/paper/Legal-Aspects-of-
Privacy-and-Security%3A-A-Case-Study/5f9139b8aa661bd040eec464d16ce31fbb829af5.
207 Andrew H. Peterson, Jesse L. Kirkpatrick, and Deborah A. Boehm-Davis, Developing Ethical,
Legal, And Policy Analyses Relevant to the Use of Machine Learning Algorithms in National Security
(white paper, George Mason University, 2017), 2, http://sites.nationalacademies.org/cs/groups/
dbassesite/documents/webpage/dbasse_179882.pdf.
67
“Some might argue that individuals consent to going outside or to other public places (i.e.,
a bank or mall) where security cameras are present.” 208

Federal laws address the use, collection, and storage of personal information. The
Driver’s Privacy Protection Act, the Gramm-Leach-Bliley Act, the Health Insurance
Portability and Accountability Act, the Fair Credit Reporting Act, and the Children’s
Online Privacy Protection Act all address personal information and were created to
establish, maintain, and protect a person’s right to privacy. These laws, however, have
severe limitations addressing the recent methods of data collection through biometric
technologies. No privacy laws exist that allow individuals to control the amount of
information that can be digitally collected without their knowledge. In addition, law
enforcement is not governed by a universal policy that would oversee the use of real-time
FRT.

Politicians are just beginning to examine the impact of FRT and the ways it affects
civil liberties and constitutional rights. The virtual absence of legislation or specific
regulations pertaining to the public, private, or government use of FRT creates a dichotomy
between law enforcement and citizens, in that trust of the police by citizens may weaken if
FRT policies are created by law enforcement without oversight. New York State has
enacted no laws pertaining to the specific use of FRT. The federal government, however,
opted to examine the feasibility of developing a framework for FRT.

On March 22, 2017, the House committee on oversight and government reform
established a subcommittee to review law enforcement’s policies on FRT, with the
intention of developing a framework for FRT, reviewing its benefits, challenges, and
various uses, and determining whether legislation is necessary. The Center on Privacy and
Technology at Georgetown Law estimated that, when accounting for all databases to which
law enforcement has access, “one in two Americans is in a facial recognition network.” 209
Law enforcement contends that a database of individuals can be beneficial whereas some

208 Ms. Smith [pseud.], “You Consent to a Search If a Camera Sees You?: Facial Recognition vs. 4th
Amendment,” CSO, March 22, 2012, https://www.csoonline.com/article/2221971/microsoft-subnet/you-
consent-to-a-search-if-a-camera-sees-you--facial-recognition-vs-4th-amendment.html.
209 Garvie, Bedoya, and Frankle, The Perpetual Line-Up, 1.

68
citizens are distrustful of law enforcement’s intentions of leveraging massive databases
created by FRT. The FBI’s Next Generation Identification System includes “an Interstate
Photo System that allows the FBI and selected state and local law enforcement to search a
database of over 30 million photos. The FBI also has agreements with at least 17 states that
allow a request for a facial recognition search of state driver’s license databases.” 210
Privacy impact assessments for the Facial Analysis, Comparison, and Evaluation Services
Unit—with the ability to compare facial images of persons associated with open
assessments and active investigations—and the Next Generation Identification–Interstate
Photo System—a mugshot repository—have been prepared by the FBI, approved by the
Department of Justice, and posted on the public-facing website of the FBI. 211 Therefore,
based on government actions, many public advocacy groups do not believe law
enforcement can effectively keep such an expansive database completely confidential.
While information sharing occurs between agencies, however, they are working to create
a transparent environment, often documenting their policies online for public consumption.

While none of these best practices are mandated, certain guidelines reflect
common-sense standards and can serve as a framework to limit liability for corporations
and law enforcement agencies. The National Telecommunications & Information
Administration has developed written guidelines for facial recognition implementers and
software operators to determine the most appropriate way to proceed in an environment
where law and policy have failed to keep pace. 212

210 Anil K. Jain, “Next Generation Biometrics” (presentation, Michigan State University, December
10, 2009), 10, http://biometrics.cse.msu.edu/Presentations/
Next_generation_biometrics_Korea_Dec2010.pdf.
211 “Privacy Impact Assessment for the Facial Analysis, Comparison, and Evaluation (FACE)
Services Unit,” Federal Bureau of Investigation, May 1, 2015, https://www.fbi.gov/services/information-
management/foipa/privacy-impact-assessments/facial-analysis-comparison-and-evaluation-face-services-
unit.
212 “Privacy Best Practice Recommendations for Commercial Facial Recognition Use,” National
Telecommunications and Information Administration, accessed March 28, 2018, https://www.ntia.doc.gov/
files/ntia/publications/
privacy_best_practices_recommendations_for_commercial_use_of_facial_recogntion.pdf.
69
 a. Data Collection/Invasiveness

Some biometric technologies, such as fingerprint scans, are embraced by law

enforcement more than other biometrics, such as FRT. Most biometric technologies gather
information with the permission or at least the knowledge of the individual. FRT, however,
can operate clandestinely and capture significant amounts of information on citizens. Real-
time FRT, if used by law enforcement, can instantly identify anonymous individuals, and
if FRT integrates with other databases, the flow of information can be unlimited. For
example, other law enforcement databases—which contain LPR, summons, and arrest data,
BWC footage, images from drones, integrated government databases, social media, open-
source data, and employment information—can provide officers with geo-coded timelines.
These can provide officers information containing both anecdotal and investigative value.
This future examination illustrates how FRT can easily integrate with other technologies,
demonstrates the capabilities of surveillance technologies, and therefore, reveals the
necessary steps for protecting and safeguarding FRT data.

b. Misuse

Critics express significant concerns that law enforcement officers may intentionally
misuse FRT for personal gain and inappropriately share confidential information. Real-
time identification of strangers, including victims of stalking or domestic violence, if used
for nefarious purposes, has chilling effects. 213 Similar to LPR technology scans when
officers use state databases to conduct name checks, officers using FRT would be able to
obtain personal data that if misused, may incur departmental and criminal sanctions.

In addition, critics raise concerns that law enforcement’s use of FRT may cause
racial bias toward minorities. Although the FBI states in its report that FRT does not
augment existing data on criminals, it still identifies minorities, African Americans in
particular, at higher rates than other races. 214 FRT can be used in law enforcement with
any device equipped with a camera. BWCs, cell phones, fixed cameras, mobile cameras,

213 Erica Klarreich, “Hello, My Name Is,” Communications of the ACM 57, no. 8 (August 2014): 19,
https://doi.org/10.1145/2632040.
214 Lynch, Face Off, 2.

70
and even drone technology can be equipped with FRT software. The methodology behind
law enforcement’s FRT data collection, sharing, and duration of storage depends on
policies created by each agency.

c. Ineffective Policies

Law enforcement organizations require transparency to ease tensions among

citizens, and therefore, when government agencies use surveillance technologies, it is
imperative to be as transparent as possible. Sabrina A. Lochner contends the use of FRT is
a serious concern if law enforcement agencies fail to inform citizens of its use. 215 Her 2012
article raises policy and privacy questions about mobile FRT and iris scans used by law
enforcement. By not making citizens aware that their personal information is being
captured, law enforcement and the private sector are putting public trust and transparency
at risk.

Another challenge for law enforcement lies in creating and establishing

comprehensive mechanisms to prevent officer misuse of FRT. As discussed previously, the
Metropolitan Police used real-time facial recognition during a November 2017
remembrance ceremony attended by 10,000 people. The department compared the faces of
individuals in the crowd to a database of approximately 50 individuals known to exhibit
dangerous behavior to public officials. 216 Martha Spurrier, director of advocacy group
Liberty, states, “There is no legal basis and no public consent for deploying this intrusive
and intimidating biometric surveillance in public space.” 217 American critics of FRT
contend that it will be similarly misused by U.S. police departments. The ACLU has shown
concern that BWCs—equipment for which the organization previously advocated to
increase police accountability—will be used as surveillance machines once integrated with
FRT. Cultural barriers to FRT, specifically its use by law enforcement, need to be

215 Sabrina A Lochner, “Saving Face: Regulating Law Enforcement’s Use of Mobile Facial
Recognition Technology & Iris Scans,” Arizona Law Review 55 (2013): 33, 202, http://www.
arizonalawreview.org/pdf/55-1/55arizlrev201.pdf.
216 Townsend, “Police to Use Facial-Recognition Cameras.”
217 Townsend.

71
addressed by examining other law enforcement technologies that are also imperfect, to
mitigate citizen privacy concerns.

Concerns over lack of transparency are not only aired in the public sector but in the
private sector as well. Facebook could be fined billions of dollars due to its facial
recognition application, which according to a class action lawsuit, violates the Illinois
Biometric Information Privacy Act. 218 Judge Donato ruled that the Illinois law is clear and
that Facebook collected a “wealth of data on its users, including self-reported residency
and IP addresses.” 219

d. Data Storage, Data Sharing, and Hacking

With the evolution of cloud technology and the ability to store large amounts of
data, law enforcement organizations are challenged with risks to stored information that
may be compromised. Each piece of collected information, if compromised, may be used
nefariously. Government databases are often the biggest target for hackers, facing daily
cyber threats that are unique because of potential harm that stolen information inflicts on
citizens. For example, the 2014 breach at the U.S. Office of Personnel Management
(OPM), in which 22 million personnel records containing personal data—work history,
family members’ names, fingerprints, and personal references—were compromised. 220
Incidents such as the OPM hacking serve as examples of how public trust in the
government erodes when a database, law enforcement or otherwise, falls prey to a security
breach.

Many civil liberty groups criticize biometric data storage and data sharing. The
ACLU urges “caution in its [biometric technology’s] deployment and stringent safeguards
in its use. 221 Use of biometrics, however, has helped the Australian Government

218 Lucas Nolan, “Facebook Could Face Billions in Fines over Facial Recognition Features,”
Breitbart, April 17, 2018, http://www.breitbart.com/tech/2018/04/17/facebook-could-face-billions-in-fines-
over-facial-recognition-features/.
219 Nolan.
220 Eggers, Delivering on Digital, loc. 2957.
221 “Biometrics,” American Civil Liberties Union, accessed July 6, 2018, https://www.aclu.org/issues/
privacy-technology/surveillance-technologies/biometrics.
72
Department of Immigration and Border Protection combat identity fraud. The international
information-sharing agreements between other countries, which involve the sharing of
facial images, allow each country to track biographical data, copies of immigration
documents, immigration status, and criminal history relevant to immigration purposes.

2. Accuracy

The NYPD will need to address accuracy concerns associated with the use of real-
time FRT, even though imperfections exist and are expected to occur with any technology.
Joy Buolamwini expresses concerns about disparities in FRT’s ability to identify races and
gender successfully. 222 Buolamwini examined the programming used in facial recognition
and concluded that machine-learning algorithms have difficulties in identifying and even
discriminate against certain races as well as gender. Her study examined three commercial
facial recognition systems and determined that darker-skinned females were most often
misclassified (34.7 percent of the time) whereas the error rate for light-skinned males was
0.8 percent. Based on Buolamwini’s research, someone may be wrongfully accused of a
crime based on erroneous yet confident misidentification of the perpetrator from security
video footage analysis.

Further adding to reliability concerns, specifically in how results may affect court
verdicts, is the problem of FRT encountering identical twins. A 2011 study contends that
FRT can identify identical twins with 90 percent accuracy. 223 More recent studies
recommend further research to determine the true scope of FRT’s ability to differentiate
between identical twins. 224 Apple, regarding its FaceID software, acknowledges that the
statistical probability is “different” for twins than for random persons, which has an

222 Joy Buolamwini and Timnit Gebru, “Gender Shades: Intersectional Accuracy Disparities in
Commercial Gender Classification,” Proceedings of Machine Learning Research 21 (2018): 1–2,
http://proceedings.mlr.press/v81/buolamwini18a/buolamwini18a.pdf.
223 P. Jonathan Phillips et al., “Distinguishing Identical Twins by Face Recognition,” in Face and
Gesture (Piscataway, NJ: IEEE, 2011), 185–187, https://doi.org/10.1109/FG.2011.5771395.
224 Raghavendra Ramachandra and Christoph Busch, “Presentation Attack Detection Methods for
Face Recognition Systems: A Comprehensive Survey,” ACM Computing Surveys 50, no. 1 (March 20,
2017): 29, https://doi.org/10.1145/3038924.
73
accuracy rate of 99.997 percent. 225 An inaccurate FRT system may lead to persons being
unnecessarily stopped by police and, thus, imposes the onus of proving one’s identity.

a. False Positives and False Negatives

Although FRT has experienced significant improvements and is more accurate than
ever before, false positives and false negatives still occur. A false positive occurs when
FRT incorrectly matches an individual to a person contained in the database. 226
Conversely, a false negative occurs when an individual is not matched, but the individual
should have been identified by the image contained within the system. 227 While false
negatives pose an unnecessary safety risk to citizens, false positives damage police
relations with the communities they serve. Stopping individuals positively identified by
facial recognition systems in crowded New York City subway stations due to false positive
matches can have damaging effects on the NYPD. In addition, the potential exists for
increased civilian complaints against officers, as well as the potential for adversarial
interactions between citizens and the police.

b. Intentional Circumvention

The potential exists for certain individuals to evade FRT cameras. Since the
inception of CCTV, those who wish to elude identification do so by simply avoiding the
cameras or covering their faces. With FRT, covering a portion of the face, wearing
sunglasses or masks, or using cosmetics to change appearance will more than likely
circumvent FRT. But as real-time FRT evolves and deep-learning neural networks continue
to advance, sunglasses and partial face covering may not outwit facial recognition software
in the future.

To further complicate the technological challenges of FRT, companies are

spawning new ideas to protect individual anonymity. Technologist Adam Harvey created

225 “About Face ID Advanced Technology,” Apple Support, accessed April 21, 2018, https://support.
apple.com/en-us/HT208108.
226 Lynch, Face Off, 6.
227 Lynch, 6.

74
a clothing line that has coded patterns designed to look fashionable, but images disguised
within the pattern confuse FRT systems and create thousands of false negatives because
current real-time facial recognition systems are unable to discern which images are real. 228

3. Transparency and Public Trust

Public support of the NYPD has transformed in recent years given the department’s
focus on neighborhood policing and strong police–community relations. The introduction
and acceptance of new surveillance technologies, however, may bring substantial criticism
that overshadows those supportive of real-time FRT.

Public trust remains a challenge for law enforcement. In addition, law enforcement
agencies wishing to pursue FRT will face additional hurdles. Matt Wood, general manager
of artificial intelligence at Amazon Web Services, points out that Amazon’s FRT has
benefited society through law enforcement’s use in preventing human trafficking,
inhibiting child exploitation, and reuniting missing children with their families. Wood
suggests there are positives and negatives with any new technology, but imposing a ban—
such as the one requested by other Amazon employees for the sale of FRT to law
enforcement—is short sighted. Society simply needs to ensure that this technology is used
the right way. 229

a. Transparency

Law enforcement organizations looking to become more transparent have

prioritized technologies that address community relations, such as BWCs, over other
technologies, regardless of their documented success in fighting crime or combating
terrorism. This is evident in the tremendous cost of BWC initiatives undertaken by law
enforcement agencies around the world. In 2016, President Obama allocated $263 million

228 Alex Hern, “Anti-Surveillance Clothing Aims to Hide Wearers from Facial Recognition,”
Guardian, January 4, 2017, http://www.theguardian.com/technology/2017/jan/04/anti-surveillance-
clothing-facial-recognition-hyperface.
229 Matt Wood, “Some Quick Thoughts on the Public Discussion Regarding Facial Recognition and
Amazon Rekognition This Past Week,” AWS New Blog, June 1, 2018, https://aws.amazon.com/blogs/aws/
some-quick-thoughts-on-the-public-discussion-regarding-facial-recognition-and-amazon-rekognition-this-
past-week/.
75
to fund 50,000 BWCs across the United States, based on his belief that more transparency
between law enforcement and citizens is needed. 230 As a result, questions and concerns of
law enforcement transparency for surveillance technologies, such as LPRs and FRT, are
not the primary focus for advocacy groups and those critical of law enforcement tactics
and strategies. In recent years, officer use-of-force encounters and police transparency have
garnered strong scrutiny in the wake of several high-profile incidents involving police and
citizens.

b. Community Mistrust of Police

Employees of major technology companies criticize company policies that allow

sharing of FRT with law enforcement. On June 21, 2018, more than 100 Amazon workers
signed and sent a letter to CEO Jeff Bezos asking the company to stop selling facial
recognition software to law enforcement. 231 The letter revealed their distrust of the police,
stating that this technology would “ultimately harm the most marginalized” based on law
enforcement’s “historic militarization of police, renewed targeting of Black activists, and
the growth of a federal deportation force currently engaged in human rights abuses.” 232
This letter came after the ACLU reported on Amazon providing facial recognition software
to law enforcement.

Citizens have different perspectives when asked about how law enforcement uses
LPR technology, BWCs, and FRT. All share similar opinions on accuracy, reliability, and
use, which can substantially affect police relations with the community. FRT, however, is
culturally perceived by critics as a technology that will do more harm than good, when
used by law enforcement. Therefore, the NYPD’s use of real-time FRT may initially
experience mixed reactions from community members.

230 Mark Landler, “Obama Offers New Standards on Police Gear in Wake of Ferguson Protests,” New
York Times, December 1, 2014, https://www.nytimes.com/2014/12/02/us/politics/obama-to-toughen-
standards-on-police-use-of-military-gear.html.
231 Greg Sandoval, “Over 100 Amazon Employees, Including Senior Software Engineers, Signed a
Letter Asking Jeff Bezos to Stop Selling Facial-Recognition Software to Police,” Business Insider, June 22,
2018, http://www.businessinsider.com/over-100-amazon-employees-sign-letter-jeff-bezos-stop-selling-
facial-recognition-software-police-2018-6.
232 Sandoval.

76
D. CONCLUSION

FRT, specifically its use in real time, serves a public benefit for citizens as well as
personal, corporate, and law enforcement use. FRT’s benefits to law enforcement are
similar to those of LPR technology as long as law enforcement and government
organizations create effective policies and ethical guidelines that govern its use. Significant
public benefits exist in the form of safer streets, and even if the probability of identifying
a known terrorist through real-time FRT turns out to be small, evidence shows that in a
controlled environment, such as the New York City subway system, this software is likely
to be successful. Additionally, the NYPD is responsible for preventing violent acts. Real-
time FRT could become an additional investigative tool that prevents further violence. For
example, in June 2018, police identified Jerrod Ramos using FRT after he walked into the
Capital Gazette in Maryland and killed five journalists. 233 The suspect was highly
uncooperative, and police needed an immediate identification to ensure no additional
threats existed.

Although little research has examined the results and effectiveness of real-time FRT
for law enforcement, direct comparisons to LPRs illustrate the similarities of benefits and
challenges in both technologies. In addition, both technologies have experienced
significant advancements yielding additional benefits and new challenges, which have
been identified by both advocacy groups and citizens. It is possible that FRT will show
similar successes and public acceptance to those of LPR technology if law enforcement
closely applies the lessons learned during LPR implementation.

233 Cade Metz and Natasha Singer, “Newspaper Shooting Shows Widening Use of Facial Recognition
by Authorities,” New York Times, July 1, 2018, https://www.nytimes.com/2018/06/29/business/newspaper-
shooting-facial-recognition.html.
77
THIS PAGE INTENTIONALLY LEFT BLANK

78
 IV. CONCLUSION

The United States is better organized and equipped to combat terrorism,

but its citizens remain fearful. The United States’ frightened, angry, and
divided society remains the country’s biggest vulnerability.

—Brian Jenkins, “Where Are We in the ‘War on Terror?” 234

Overall, this thesis concludes that, while decision-makers must mitigate challenges
to receive public support, real-time FRT is not nearly as frightening as it might seem to
some. In other words, having compared it to LPR technology, this thesis lays out the tools
law enforcement leaders already have to mitigate the fears, particularly in connection with
well-established policies and practices that protect civil liberties. Once law enforcement
mitigates the challenges of FRT, the benefits will surely outweigh them. Law enforcement
use of FRT must be ethical, transparent, and responsible. Other places, such as China, have
different visions of its intended use, which do not conform to the values and privacy
considerations upheld by the NYPD. This chapter concludes the thesis by reviewing the
fundamental challenges and benefits identified with LPRs as well as the lessons learned
from BWCs, so the NYPD can apply them in evaluating real-time FRT. The chapter ends
with thoughts on why the NYPD should mitigate the challenges and gain public support,
so the community as a whole may benefit from the real-time use of FRT.

LPR technology is not a panacea for solving all crimes, countering all terrorist
actions, or finding every lost child, but LPRs provide law enforcement with significant
benefits. Additionally, LPRs can be integrated with other law enforcement technologies.
The capabilities of LPR technology have recently expanded, with new uses for both law
enforcement and the private sector, despite the concerns identified in this thesis. Benefits
relating to crime prevention illustrate the growing trend in which LPR technology uses new
methods to enforce the laws, such as red light and speed cameras to enforce traffic codes.

234 Brian M. Jenkins, “Fifteen Years on, Where Are We in the ‘War on Terror?,’” Combating
Terrorism Center, September 7, 2016, https://ctc.usma.edu/fifteen-years-on-where-are-we-in-the-war-on-
terror/.
79
The expansion of LPR technology in New York City demonstrates that citizens are not
overtly opposed to its use.

The NYPD has developed sound policies regarding LPRs and already mitigated
many of the challenges to minimize potential harm to civil liberties and misuse. In addition,
LPRs are generally accepted as a common law enforcement technology, largely in part
because the NYPD previously addressed many of the risks. The NYPD addressed LPR data
collection, retention, and sharing through robust and clear policies. In addition, the
department addressed potential misuse of LPRs to eliminate the ambiguities and clearly
define acceptable practices. The NYPD can apply the same strategies to address the
challenges with FRT because both technologies perform similar functions. FRT and LPRs
are alike in that both technologies collect all available information, without bias, from
license plates or persons. Furthermore, both technologies use hot lists to compare against
collected data, and the information received requires human intervention to verify the
information, such as a stolen vehicle or a wanted person.

A police officer with a sharp eye can detect a stolen vehicle from information
contained in a police report, which is similar to a police officer recognizing the face of an
individual wanted for a previous crime. LPRs and real-time FRT both act as force
multipliers for police officers on the street by scanning vehicles and persons. In New York
City, a police officer could not possibly do all the work that technology can do, and
therefore, technology can assist yet not replace police officers. In the NYPD, public
acceptance of LPR exists because many of the challenges have been addressed. Therefore,
the NYPD can apply a similar framework to real-time FRT whereby the department already
addressed LPRs.

The NYPD should consider real-time FRT to help facilitate greater efficiencies in
policing and to work in conjunction with other technologies and law enforcement tools. In
addition, real-time FRT can be used by the NYPD as another tool for police officers in the
field. For example, NYPD police officers carry many less-lethal devices—pepper spray,
conducted electrical devices, and impact weapons—so if one tool is ineffective in a given
situation, officers have other options to mitigate a potential threat. Real-time FRT can be

80
used alongside other technologies and equipment to enhance public safety and maintain
law and order.

Further research will be necessary to determine the overall effectiveness of actual

real-time FRT. The benefits of integrating real-time FRT with other technology platforms,
such as smartphones and BWCs, could be immeasurable. The challenges, however, must
be carefully managed to successfully implement a system that garners public acceptance
by the majority. Overall, negative connotations surrounding the use of real-time FRT,
specifically when used by law enforcement, have emerged because the public benefit has
not yet been established. With LPRs, documented evidence shows their benefit as a force
multiplier in reducing crime, aiding in investigations, conducting summons enforcement,
easing traffic congestion, allowing toll payments, and identifying missing persons, wanted
felons, and potential terrorists. The challenges faced by law enforcement—privacy
concerns, officer misuse, inaccuracy, lack of transparency, as well as the methods by which
data is collected, stored, and shared—are not entirely solved, but in New York City, current
policies and procedures ease many of the concerns and show that citizens are not strongly
opposed to the use of LPR technology. The benefits and challenges of LPRs, FRT, and
BWCs for law enforcement are enumerated in Tables 1 and 2, respectively.

81
 Table 1. Comparison of Benefits of LPRs, FRT, and BWCs

Facial Recognition License Plate Body-Worn

Technology Readers Cameras
Effective;
Effective in the
Encompasses Effective in Targeted
Crime Prevention Wearer’s Peripheral
Crowded Areas
Vision, Surroundings
Environments
Investigating Precautionary
Remedial Measures Uncorrelated
Terrorism Measures
Instant Record of
Real-Time IDs of Instant ID of Instant ID of Plate
Wearer’s Peripheral
Wanted/Known Individuals Numbers
Vision, Surroundings
Effective; Easy to Effective; Easy to Effective in Officer
Investigative
Detect Suspects, Detect Vehicle Assessments and
Assistance
Wanted Persons History Engagements
Unbiased Policing;
Uncorrelated, Unless
Increased Mindful of One’s
Officer Safety Integrated with Other
Awareness; Safe Actions
Tech
Driving
Instant IDs of
Instant ID of Person;
Vehicle Owner’s Within Wearer’s
Mental Health Support Emergency, Special
License, Health Ability
Needs Info
Conditions
Instant ID of Amber Alerts, Within Wearer’s
Missing Persons
Individual’s Location However, Limited Ability
Eliminates
Overcrowding in Turnstiles; Utilizes
Uncorrelated Uncorrelated
Transit EZ-Pass Method
for Transit
Integration with Other
Versatile Limited Limited
Tech

82
 Table 2. Comparison of Challenges with LPRs, FRT, and BWCs

Facial Recognition License Plate Body-Worn

Technology Readers Cameras
Continuous Data
Privacy: Stores Vehicle Video Footage
Collection of
Data Collection; Records Regardless Recorded at the
Citizen’s Personal
Invasiveness of Illicit Activity Discretion of Wearer
Info
Personal Records Camera Can Be
Abuse of Vehicular
Can Be Abused; Deactivated during
Privacy: Misuse Activity; Selective
May Cause Racial Unfavorable
Enforcement
Bias Situation
Difficult to Create,
Establish
Inconsistent Policies
Privacy: Ineffective Comprehensive Policies Vary by
within Each
Policies Mechanisms to Agency
Department
Prevent Officer
Misuse
Requires Abundant
Risk of Personal
Requires Abundant Storage Space; Risk
Privacy: Data Storage, Data Being
Storage Space; Risk of Video Being
Sharing, Hacking Compromised by
of Cyber Attacks Compromised by
Hackers
Hackers
Difficulties
Accuracy: Unnecessary Safety Distinguishing
False Positives; False Risks; Damage to between Identical Uncorrelated
Negatives Police Relations Plates from Different
States
Can Avoid Cameras
Accuracy: Intentional LPR Signals Can Be Camera Can Be
or Cover Portions of
Circumvention Disrupted Deactivated
the Face
Decision to Activate
Tech Addressing Tech Addressing
Camera Based on
Community Community
Transparency Department Policy,
Relations Is Relations Is
Transparency,
Prioritized Prioritized
Individual Privacy
Connections of Footage of
Distrust of Police
Public Mistrust of Vehicle History to Misconduct Raise
Using Tech to Harm
Police That of Illicit Questions and
Marginalized Groups
Behavior Distrust

Technology is part of a crucial conversation in policy now. For example, BWCs

were seen as an effective response in holding police accountable for use-of-force
encounters as well as providing transparency in policing. The result of their use, however,
created controversies between law enforcement and citizens due to inconsistent policies
from one police department to another. BWCs were supposed to remove the ambiguity of
policing and create an environment where policing is transparent, fosters stronger relations

83
between police and community, and decreases civilian complaints against police officers
as well as use-of-force incidents. Instead, BWCs created a new level of privacy concerns;
those concerns, however, are distinct from both LPRs and FRT. Officers use BWCs in most
police interactions, often recording inside private residences and potentially capturing
moments that citizens do not want memorialized on video. LPRs and FRT, on the other
hand, record already public information—license plates and faces.

Law enforcement organizations, looking to become more transparent, have

prioritized BWC implementation over other technologies to address community relations.
U.S. law enforcement and legislators equally support the use of BWCs, but they fear law
enforcement’s use of real-time FRT will send the message that citizen privacy has been
compromised, which is simply not a valid concern. This thesis concludes that the
expectation of privacy inside a private residence greatly differs from privacy expectations
in a public setting.

To address public trust concerns and create a model that fosters transparency, the
NYPD initiated a robust strategy to rebrand the NYPD image. The strategy behind
neighborhood policing facilitates collaboration between local police officers and members
of the community to share the responsibility of safety in New York City neighborhoods,
allowing communities to work in partnership with the NYPD. Neighborhood policing,
combined with Build the Block meetings, facilitates conversations and answers questions
about various surveillance technologies, including ideas from citizens about how
communities want to be policed. The NYPD must continue to take proactive steps, using
neighborhood policing as a platform for better community relations, transparency, and the
opportunity for citizens to understand how the NYPD can and will use real-time FRT. Once
the NYPD garners public acceptance, it may realize the use of real-time FRT for a safer
New York City subway system.

84
 LIST OF REFERENCES

Allevate. “Helping to Counter the Terrorist Threat Using Face Recognition.” July 28,
2017. http://allevate.com/index.php/2017/07/28/helping_to_counter_the_terrorist_
threat_using_face_recognition/.

American Civil Liberties Union. “Biometrics.” Accessed July 6, 2018. https://www.aclu.

org/issues/privacy-technology/surveillance-technologies/biometrics.

Anderson, Lisa. “Exclusive-Poll: New York City Transport Seen as Safest in World for
Women.” Reuters. October 28, 2014. https://uk.reuters.com/article/women-poll-
newyork/exclusive-poll-new-york-city-transport-seen-as-safest-in-world-for-
women-idUKL6N0SB4WI20141029.

Apple Support. “About Face ID Advanced Technology.” Accessed April 21, 2018.
https://support.apple.com/en-us/HT208108.

Baker, Al. “License Plate Cameras Aid in Police Investigations.” New York Times, April
11, 2011. https://www.nytimes.com/2011/04/12/nyregion/12plates.html.

———. “Street Stops by New York City Police Have Plummeted.” New York Times,
May 30, 2017. https://www.nytimes.com/2017/05/30/nyregion/nypd-stop-and-
frisk.html.

Bennett, Kanya A. “Can Facial Recognition Technology Be Used to Fight the New Way
against Terrorism: Examining the Constitutionality of Facial Recognition
Surveillance Systems.” North Carolina Journal of Law & Technology 3 (2001):
151–174. http://ncjolt.org/can-facial-recognition-technology-be-used-to-fight-the-
new-war-against-terrorism-examining-the-constitutionality-of-facial-recognition-
surveillance-systems/.

Bignami, Francesca, and Giorgio Resta. Human Rights Extraterritoriality: The Right to
Privacy and National Security Surveillance. Rochester, NY: Social Science
Research Network, 2018. https://papers.ssrn.com/sol3/papers.cfm?abstract_
id=3043771.

Biometrica Systems. “Biometrica Announces Release of New SSIN, with Mobile App
That Allows Near Real-Time Facial Recognition.” November 23, 2016.
https://biometrica.com/biometrica-announces-release-of-new-ssin-with-mobile-
app-that-allows-near-real-time-facial-recognition/.

Bledsoe, Woodrow Wilson. Proposal for a Study to Determine the Feasibility of a

Simplified Facial Recognition Machine. Palo Alto, CA: Panoramic Research,
January 30, 1963. http://archive.org/details/firstfacialrecognitionresearch.

85
Bowles, Nellie. “‘I Think My Blackness Is Interfering’: Does Facial Recognition Show
Racial Bias?” Guardian, April 8, 2016. http://www.theguardian.com/technology/
2016/apr/08/facial-recognition-technology-racial-bias-police.

Brown, Steven Rex. “NYPD Ripped for Abusing Facial-Recognition Tool.” New York
Daily News, March 1, 2018. http://www.nydailynews.com/new-york/nyc-
crime/nypd-ripped-abusing-facial-recognition-tool-article-1.3847796.

Buolamwini, Joy, and Timnit Gebru. “Gender Shades: Intersectional Accuracy

Disparities in Commercial Gender Classification.” Proceedings of Machine
Learning Research 21 (2018): 1–15. http://proceedings.mlr.press/v81/
buolamwini18a/buolamwini18a.pdf.

Byrne, John. “Emanuel Raises Facial Recognition Tech in Taxi vs. Ride-Share Debate.”
Chicago Tribune, November 7, 2017. http://www.chicagotribune.com/news/
local/politics/ct-met-rahm-emanuel-rideshare-report-20171107-story.html.

Cavadini, Daniele, Andreas Meier, Daniel Fasel, and Lorenzo Cimasoni. “Introducing the
Biometrical Electronic Passport (EPass).” Seminar, University of Fribourg, 2008.

City of New York. “Vision Zero: Mayor de Blasio Pushes for Tougher State Laws to
Keep Dangerous Drivers off the Streets.” March 15, 2018. https://www1.nyc.gov/
office-of-the-mayor/news/136-18/vision-zero-mayor-de-blasio-pushes-tougher-
state-laws-keep-dangerous-drivers-off-streets#/0.

City of New York, Police Department Legal Bureau. “Freedom of Information Law
Request: #14-PL-0175.” Letter to Nate Vogel, New York Civil Liberties Union,
April 17, 2014. https://www.nyclu.org/sites/default/files/20140417_
NYPDDenial_NYCLUFOILRequest.pdf.

Communities United for Police Reform. “Know Your Rights! Help End Discriminatory,
Abusive & Illegal Policing!” Accessed May 6, 2018. http://changethenypd.org/
resources/know-your-rights-help-end-discriminatory-abusive-illegal-policing.

Constitution Project. “Law Enforcement Facial Recognition Is a Powerful Surveillance

Technology in Need of Independent Checks and Limits.” March 30, 2017.
https://constitutionproject.org/wp-content/uploads/2017/03/Facial-Recognition-
Statement-for-Record_The-Constitution-Project.pdf.

Counter Extremism Project. “Vehicles as Weapons of Terror.” May 22, 2017.

https://www.counterextremism.com/vehicles-as-weapons-of-terror.

Crump, Catherine. “An Ethical Framework for Face Recognition.” American Civil
Liberties Union. July 16, 2013. https://www.ntia.doc.gov/files/ntia/publications/
aclu_an_ethical_framework_for_face_recognition.pdf.

86
———. You Are Being Tracked: How License Plate Readers Are Being Used to Record
Americans’ Movements. New York: American Civil Liberties Union, July 2013.
https://www.aclu.org/issues/privacy-technology/location-tracking/you-are-being-
tracked.

Cuador, Claudia. “From Street Photography to Face Recognition: Distinguishing between

the Right to Be Seen and the Right to Be Recognized.” Nova Law Review 41, no.
2 (2017): 1–28. https://nsuworks.nova.edu/cgi/viewcontent.cgi?referer=
https://scholar.google.com/&httpsredir=1&article=1998&context=nlr/.

Customs and Border Protection. “CBP Deploys Facial Recognition Biometric

Technology at 1 TSA Checkpoint at JFK Airport.” October 11, 2017.
https://www.cbp.gov/newsroom/national-media-release/cbp-deploys-facial-
recognition-biometric-technology-1-tsa-checkpoint.

Daniel, Stephanie, Nicole Lewis, and Kalalea Kalalea. “In the Wake of Broken Windows
Policing How Aggressive Policing Contributed to East Harlem Residents Distrust
of Police.” Master’s capstone, City University of New York, 2016.

Davis, Wendy. “Facial Recognition Technology Nab Criminals-and Raises Privacy

Concerns.” ABA Journal. October 2017. http://www.abajournal.com/magazine/
article/facial_recognition_technology_crime_privacy.

Dees, Tim. “Finding Stolen Vehicles.” Hendon Media Group. Accessed June 3, 2018.
http://www.hendonpub.com/resources/article_archive/results/details?id=3901.

Del Greco, Kimberly J. “Law Enforcement’s Use of Facial Recognition Technology.”

Federal Bureau of Investigations. March 22, 2017. https://www.fbi.gov/news/
testimony/law-enforcements-use-of-facial-recognition-technology.

Department of Homeland Security. “Privacy Impact Assessment for the Automated

Biometric Identification System (IDENT).” Accessed August 8, 2018.
https://www.dhs.gov/sites/default/files/publications/privacy-pia-nppd-
identappendices-august2017.pdf.

Deshmukh, Shubhada, Manasi Patwardhan, and Anjali Mahajan. “Survey on Real-Time

Facial Expression Recognition Techniques.” IET Biometrics 5, no. 3 (September
2016): 155–163.

Detsing, Jeerawat, and Mahasak Ketcham. “Detection and Facial Recognition for
Investigation.” In International Conference on Digital Arts, Media and
Technology, 407–411. Piscataway, NJ: IEEE, 2017. https://doi.org/10.1109/
ICDAMT.2017.7905002.

Dodd, Vikram. “Met Police to Use Facial Recognition Software at Notting Hill
Carnival.” Guardian, August 5, 2017. http://www.theguardian.com/uk-news/
2017/aug/05/met-police-facial-recognition-software-notting-hill-carnival.
87
Doubek, James. “Digital License Plates Roll Out in California.” NPR. June 1, 2018.
https://www.npr.org/sections/thetwo-way/2018/06/01/616043976/digital-license-
plates-roll-out-in-california.

Drover, Paul, and Barak Ariel. “Leading an Experiment in Police Body-Worn Video
Cameras.” International Criminal Justice Review 25, no. 1 (March 1, 2015): 80–
97. https://doi.org/10.1177/1057567715574374.

Du, Shan, M. Ibrahim, M. Shehata, and W. Badawy. “Automatic License Plate

Recognition: A State-of-the-Art Review.” IEEE Transactions on Circuits and
Systems for Video Technology 23, no. 2 (February 2013): 311–325.
https://doi.org/10.1109/TCSVT.2012.2203741.

Eggers, William D. Delivering on Digital: The Innovators and Technologies That Are
Transforming Government. 1st ed. New York: Rosetta Books, 2016.

Eisler, Ben. “ACLU Concerned Automatic License Plate Readers May Invade Privacy.”
WJLA News. July 30, 2012. http://wjla.com/news/local/aclu-concerned-red-light-
cameras-may-invade-privacy-78301.

Federal Bureau of Investigation. “2015 Crime in the United States.” Accessed June 3,
2018. https://ucr.fbi.gov/crime-in-the-u.s/2015/crime-in-the-u.s.-
2015/tables/table-1.

———. “2017 NCIC Missing Person and Unidentified Person Statistics.” Accessed June
17, 2018. https://www.fbi.gov/file-repository/2017-ncic-missing-person-and-
unidentified-person-statistics.pdf/view.

———. “Privacy Impact Assessment for the Facial Analysis, Comparison, and
Evaluation (FACE) Services Unit.” May 1, 2015. https://www.fbi.gov/services/
information-management/foipa/privacy-impact-assessments/facial-analysis-
comparison-and-evaluation-face-services-unit.

Find Biometrics. “Facial Recognition.” Accessed August 7, 2018.

https://findbiometrics.com/solutions/facial-recognition/.

Fitzsimmons, Emma G. “Every New York City Subway Line Is Getting Worse. Here’s
Why.” New York Times, June 28, 2017. https://www.nytimes.com/interactive/
2017/06/28/nyregion/subway-delays-overcrowding.html.

Fortune. “Full Transcript of Hillary Clinton’s NAACP Speech: ‘This Madness Has to
Stop.’” July 18, 2016. http://fortune.com/2016/07/18/hillary-clinton-speech-
naacp-transcript/.

88
Francescani, Chris. “NYPD Expands Surveillance Net to Fight Crime as well as
Terrorism.” Reuters. June 21, 2013. https://www.reuters.com/article/usa-ny-
surveillance/nypd-expands-surveillance-net-to-fight-crime-as-well-as-terrorism-
idUSL2N0EV0D220130621.

Future of Privacy Forum. Privacy Principles for Facial Recognition Technology.

Washington, DC: Future of Privacy Forum, December 2015. https://fpf.org/wp-
content/uploads/2015/12/Dec9Working-Paper-
FacialRecognitionPrivacyPrinciples-For-Web.pdf.

Garvie, Clare, Alvaro Bedoya, and Johnathan Frankle. The Perpetual Line-Up:
Unregulated Police Face Recognition in America. Washington, DC: Center on
Privacy and Technology, Georgetown Law, October 18, 2016. https://www.
perpetuallineup.org.

Gates, Kelly. “Identifying the 9/11 ‘Faces of Terror.’” Cultural Studies 20, no. 4/5
(September 7, 2006): 417–440. https://doi.org/10.1080/09502380600708820.

Gavison, Ruth. “Privacy and the Limits of Law.” Yale Law Journal 89, no. 3 (1980):
421–471. https://www.jstor.org/stable/pdf/795891.pdf.

Gierlack, Keith, Shara Williams, Tom LaTourrette, James M. Anderson, Lauren A.

Mayer, and Johanna Zmud. License Plate Readers for Law Enforcement:
Opportunities and Obstacles. Santa Monica, CA: RAND Corporation, 2014.
https://www.rand.org/pubs/research_reports/RR467.html.

Goodman, J. David. “Court-Appointed Police Monitor Has Fought for City and against
It.” New York Times, August 13, 2013. https://www.nytimes.com/2013/08/14/
nyregion/court-appointed-police-monitor-has-fought-for-city-and-against-it.html.

Goodman, J. David, and Al Baker. “Police Shoot Hammer-Wielding Man Sought in 4

Manhattan Attacks.” New York Times, May 13, 2015. https://www.nytimes.com/
2015/05/14/nyregion/officer-shoots-man-in-midtown-manhattan.html.

Gore, Lynn. “Request for Information (RFI), RFI-16-63, All-Electronic Facial Detection
and Recognition System at all TBTA Facilities.” Official memorandum, MTA
Bridges and Tunnels, December 12, 2016. https://www.documentcloud.org/
documents/3428597-RFI.html.

Government Accountability Office. Facial Recognition Technology: Commercial Uses,

Privacy Issues, and Applicable Federal Law. GAO-15-621. Washington, DC:
GAO, July 2015.

Griffin, Joel. “LPR Technology Comes of Age.” Security Info Watch. August 4, 2015.
http://www.securityinfowatch.com/article/12099279/lpr-technology-comes-of-
age.

89
Guzzardo, Jamie, and Jesse Solomon. “Missing Boy Spent 11 Days Wandering New
York Subways.” CNN. November 25, 2009. http://www.cnn.com/2009/US/
11/25/new.york.subway.teen/index.html.

Haigney, Sophie. “ACLU Sues Immigration and Customs Enforcement for License Plate
Reader Records.” SF Gate. May 28, 2018. https://www.sfgate.com/bayarea/
article/ACLU-sues-ICE-for-license-plate-reader-contracts-12937712.php.

Hern, Alex. “Anti-Surveillance Clothing Aims to Hide Wearers from Facial

Recognition.” Guardian, January 4, 2017. http://www.theguardian.com/
technology/2017/jan/04/anti-surveillance-clothing-facial-recognition-hyperface.

Higginbotham, Stacy. “Facial Recognition Freak Out: What the Technology Can and
Can’t Do.” Fortune, June 23, 2015. http://fortune.com/2015/06/23/facial-
recognition-freak-out/.

Hirose, Mariko. “Documents Uncover NYPD’s Vast License Plate Reader Database.”
American Civil Liberties Union. January 25, 2016. https://www.aclu.org/blog/
privacy-technology/location-tracking/documents-uncover-nypds-vast-license-
plate-reader-database.

Hsu, Hwai-Jung, and Kuan-Ta Chen. “Face Recognition on Drones: Issues and
Limitations.” In DroNet ‘15: Proceedings of the First Workshop on Micro Aerial
Vehicle Networks, Systems, and Applications for Civilian Use, 39–44. New York:
ACM Press, 2015. http://dl.acm.org/citation.cfm?doid=2750675.2750679.

Huggler, Justin. “Facial Recognition Software to Catch Terrorists Being Tested at Berlin
Station.” Telegraph, August 2, 2017. http://www.telegraph.co.uk/news/2017/08/
02/facial-recognition-software-catch-terrorists-tested-berlin-station/.

INTERPOL. “Forensics.” Accessed July 3, 2018. https://www.interpol.int/INTERPOL-

expertise/Forensics/Facial-recognition.

Introna, Lucas, and Helen Nissenbaum. Facial Recognition Technology: A Survey of

Policy and Implementation Issues. New York: Center for Catastrophe
Preparedness and Response, 2010. https://www.nyu.edu/projects/nissenbaum/
papers/facial_recognition_report.pdf.

Ito, Koichi, and Takafumi Aoki. “Recent Advances in Biometric Recognition.” ITE
Transactions on Media Technology and Applications 6, no. 1 (2018): 64–80.
https://doi.org/10.3169/mta.6.64.

ITV News. “Facial Recognition Technology ‘Could Prevent Future Terror Attacks.’”
March 1, 2018. http://www.itv.com/news/london/2018-03-01/facial-recognition-
technology-could-prevent-future-terror-attacks/.

90
Jain, Anil K. “Next Generation Biometrics.” Presentation, Michigan State University,
December 10, 2009. http://biometrics.cse.msu.edu/Presentations/Next_
generation_biometrics_Korea_Dec2010.pdf.

Jain, Surbhi. “Big Data: What, Why and Why Not.” International Journal of Engineering
Development 5, no. 2 (2017). https://www.ijedr.org/papers/IJEDR1702334.pdf.

Jeberson, W., and Lucky Sharma. “Survey on Big Data for Counter Terrorism.”
International Journal of Innovations and Advancement in Computer Science 4
(May 2015): 197–205.

Jenkins, Brian M. “Fifteen Years on, Where Are We in the ‘War on Terror?’” Combating
Terrorism Center. September 7, 2016. https://ctc.usma.edu/fifteen-years-on-
where-are-we-in-the-war-on-terror/.

Kapp, Trevor. “Neighborhood Policing Changing Attitudes and Reaping Benefits, NYPD
Says.” DNA Info. April 6, 2017. https://www.dnainfo.com/new-york/20170406/
central-harlem/nypd-neighborhood-policing-nco-community-police-department.

Kilgannon, Corey, and Joseph Goldstein. “Sayfullo Saipov, the Suspect in the New York
Terror Attack, and His Past.” New York Times, November 1, 2017.
https://www.nytimes.com/2017/10/31/nyregion/sayfullo-saipov-manhattan-truck-
attack.html.

King, Rawlson O’Neil. Special Report: Biometrics in Law Enforcement. Lansing, MI:
Biometrics Research Group, Michigan State University, 2017. https://www.
biometricupdate.com/wp-content/uploads/2017/08/special-report-biometrics-in-
law-enforcement.pdf.

Koen, Marthinus C., James J. Willis, and Stephen D. Mastrofski. “The Effects of Body-
Worn Cameras on Police Organisation and Practice: A Theory-Based Analysis.”
Policing and Society (April 2018): 1–17. https://www.tandfonline.com/doi/pdf/
10.1080/10439463.2018.1467907?needAccess=true.

Kolkey, Jeff. “How One Illinois City Uses Automatic License Plate Readers and Other
Police Tech.” Government Technology. May 21, 2018. http://www.govtech.com/
dc/How-One-Illinois-City-Uses-Automatic-License-Plate-Readers-and-Other-
Police-Tech.html.

Koper, Christopher S., Bruce G. Taylor, and Daniel J. Woods. “A Randomized Test of
Initial and Residual Deterrence from Directed Patrols and Use of License Plate
Readers at Crime Hot Spots.” Journal of Experimental Criminology 9, no. 2 (June
2013): 213–244. https://www.voiceofsandiego.org/wp-content/uploads/2018/04/
Koper-et-al-2.pdf.

Klarreich, Erica. “Hello, My Name Is.” Communications of the ACM 57, no. 8 (August
2014): 17–19. https://doi.org/10.1145/2632040.
91
Kravets, David. “Enhanced DMV Facial Recognition Technology Helps NY Nab 100 ID
Thieves.” Ars Technica, August 28, 2016. https://arstechnica.com/tech-
policy/2016/08/enhanced-dmv-facial-recognition-technology-helps-ny-nab-100-
id-thieves/.

Landler, Mark. “Obama Offers New Standards on Police Gear in Wake of Ferguson
Protests.” New York Times, December 1, 2014. https://www.nytimes.com/
2014/12/02/us/politics/obama-to-toughen-standards-on-police-use-of-military-
gear.html.

Lane, Julia, Victoria Stodden, Stefan Bender, and Helen Nissenbaum. Privacy, Big Data,
and the Public Good: Frameworks for Engagement. Cambridge: Cambridge
University Press, 2014. https://www.nyu.edu/projects/nissenbaum/papers/
BigDatasEndRun.pdf.

Laperruque, Jake. “Preserving the Right to Obscurity in the Age of Facial Recognition.”
Century Foundation. October 20, 2017. https://tcf.org/content/report/preserving-
right-obscurity-age-facial-recognition/.

Lochner, Sabrina A. “Saving Face: Regulating Law Enforcement’s Use of Mobile Facial
Recognition Technology & Iris Scans.” Arizona Law Review 55 (2013): 202–233.
http://www.arizonalawreview.org/pdf/55-1/55arizlrev201.pdf.

Lum, Cynthia, Julie Hibdon, Breanne Cave, Christopher S. Koper, and Linda Merola.
“License Plate Reader (LPR) Police Patrols in Crime Hot Spots: An Experimental
Evaluation in Two Adjacent Jurisdictions.” Journal of Experimental Criminology
7, no. 4 (2011): 321–345. https://www.voiceofsandiego.org/wp-content/uploads/
2018/04/Lum-et-al-2.pdf.

Lynch, Jennifer. Face Off, Law Enforcement Use of Face Recognition Technology. San
Francisco: Electronic Frontier Foundation, February 12, 2018. https://www.eff.
org/wp/law-enforcement-use-face-recognition.

Maass, Dave. “The Four Flavors of Automated License Plate Reader Technology.”
Electronic Frontier Foundation. April 6, 2017. https://www.eff.org/deeplinks/
2017/04/four-flavors-automated-license-plate-reader-technology.

Maleske, Melissa. “Facial Recognition Presents Privacy Concerns.” Inside Counsel.

March 2012. https://www.law.com/almID/4f46a3f8150ba0c00a000048/.

MAS Context. “Ring of Steel.” June 25, 2014. http://www.mascontext.com/issues/22-

surveillance-summer-14/ring-of-steel/.

McClain, Noah. “The Horizons of Technological Control: Automated Surveillance in the

New York Subway.” Information, Communication & Society 21, no. 1 (2018):
46–62. https://doi.org/10.1080/1369118X.2016.1260624.

92
McCoy, Susan. “O’ Big Brother Where Art Thou?: The Constitutional Use of Facial-
Recognition Technology.” John Marshall Journal of Computer & Information
Law 20, no. 3 (Spring 2002): 471–494.

Mell, Peter, and Timothy Grance. The NIST Definition of Cloud Computing:
Recommendations of the National Institute of Standards and Technology. Special
Publication 800–145. Gaithersburg, MD: NIST, September 2011. http://nvlpubs.
nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-145.pdf

Melodia, Mark, Paul Bond, and Angela-Angelovska-Wilson. “Legal Risks and Rules of
the Move to Biometrics.” New York Law Journal, March 2, 2015. https://www.
technologylawdispatch.com/wp-content/uploads/sites/26/2016/02/Legal-NYLJ-
Article-Risks-and-Rules-of-the-Move-to-Biometrics.pdf.

Merola, Linda M., and Cynthia Lum. “Understanding Citizen Support for License Plate
Readers.” Translational Criminology (2015): 1–29. https://www.bja.gov/bwc/
pdfs/TC8-Spring2015.pdf#page=25.

Metropolitan Transportation Authority. “Introduction to Subway Ridership.” Accessed

September 4, 2017. http://web.mta.info/nyct/facts/ridership/.

Metz, Cade, and Natasha Singer. “Newspaper Shooting Shows Widening Use of Facial
Recognition by Authorities.” New York Times, July 1, 2018. https://www.nytimes.
com/2018/06/29/business/newspaper-shooting-facial-recognition.html.

Miller, Lindsay, and Jessica Toliver. Implementing a Body-Worn Camera Program:

Recommendations and Lessons Learned. Washington, DC: Community Oriented
Policing Services, 2014. https://www.justice.gov/iso/opa/resources/
472014912134715246869.pdf.

Möllers, Norma, and Jens Hälterlein. “Privacy Issues in Public Discourse: The Case of
‘Smart’ CCTV in Germany.” Innovation: European Journal of Social Sciences 26
(March 1, 2013): 65–87. https://publishup.uni-potsdam.de/opus4-ubp/frontdoor/
deliver/index/docId/7767/file/moellers_diss.pdf#page=73.

Moriarty, Laura J., ed. Criminal Justice Technology in the 21st Century. 3rd ed.
Springfield, IL: Charles C. Thomas Publisher, 2017.

Moy, Laura M., Harrison Rudolph, and Alvaro M. Bedoya. Not Ready for Takeoff: Face
Scans at Airport Departure Gates. Washington, DC: Center on Privacy and
Technology, Georgetown Law, December 21, 2017. https://www.
airportfacescans.com/sites/default/files/Biometrics_Report__Not_Ready_For_Tak
eoff.pdf

Moynihan, Colin. “Driver Charged with Manslaughter in Deaths of 2 Children.” New

York Times, May 3, 2018. https://www.nytimes.com/2018/05/03/nyregion/driver-
manslaughter-brooklyn-children.html.
93
Mozur, Paul. “Inside China’s Dystopian Dreams: A.I., Shame and Lots of Cameras.” New
York Times, July 9, 2018. https://www.nytimes.com/2018/07/08/business/china-
edsurveillance-technology.html.

Ms. Smith [pseud.]. “You Consent to a Search If a Camera Sees You?: Facial
Recognition vs. 4th Amendment.” CSO. March 22, 2012. https://www.csoonline.
com/article/2221971/microsoft-subnet/you-consent-to-a-search-if-a-camera-sees-
you--facial-recognition-vs-4th-amendment.html.

Mullins, J. “Ring of Steel II: New York City Gets Set to Replicate London’s High-
Security Zone.” IEEE Spectrum 43, no. 7 (July 2006): 12–13. https://ieeexplore.
ieee.org/stamp/stamp.jsp?tp=&arnumber=1652996.

Munger, Kristen, and Shelby J. Harris. “Effects of an Observer on Handwashing in a

Public Restroom.” Perceptual and Motor Skills 69 (1989): 733–734.
https://kmunger.files.wordpress.com/2007/09/munger-harris-1989-effects-of-an-
observer-on-handwashing-in-a-public-restroom.pdf.

National Alliance on Mental Illness. “Mental Health by the Numbers.” Accessed June 17,
2018. https://www.nami.org/Learn-More/Mental-Health-By-the-Numbers.

National Conference of State Legislatures. “Automated License Plate Readers: State

Legislation 2016 & 2017.” Accessed May 6, 2018. http://www.ncsl.org/research/
telecommunications-and-information-technology/automated-license-plate-
readers-state-legislation-2016.aspx.

National Law Enforcement Officers Memorial Fund. 2017 End of Year Officer Fatalities
Report. Washington, DC: NLEOMF, 2017. http://www.nleomf.org/assets/pdfs/
reports/fatality-reports/2017/2017-End-of-Year-Officer-Fatalities-
Report_FINAL.pdf.

National Telecommunications and Information Administration. “Privacy Best Practice

Recommendations for Commercial Facial Recognition Use.” Accessed March 28,
2018. https://www.ntia.doc.gov/files/ntia/publications/privacy_best_practices_
recommendations_for_commercial_use_of_facial_recogntion.pdf.

NEC Global Face Recognition Centre of Excellence. NeoFace Watch, High Performance
Face Recognition. NEC Global Face Recognition Centre of Excellence, 2016.
https://www.nec.com/en/global/solutions/safety/face_recognition/PDF/Face_Reco
gnition_NeoFace_Watch_Brochure.pdf.

———. “Technology.” Accessed April 15, 2018, http://www.nec.com/en/global/

solutions/safety/Technology/FaceRecognition/index.html.

Newell, Bryce Clayton. “Local Law Enforcement Jumps on the Big Data Bandwagon:
Automated License Plate Recognition Systems, Information Privacy, and Access
to Government Information.” Maine Law Review 66 (2013): 1–42.
94
New York City Police Department. “Build the Block.” Accessed July 17, 2018.
https://www1.nyc.gov/site/nypd/bureaus/patrol/buildtheblock.page.

———. “Counterterrorism.” Accessed May 14, 2018. https://www1.nyc.gov/site/nypd/

bureaus/investigative/counterterrorism.page.

———. “Crime Statistics.” Accessed July 17, 2018. https://www1.nyc.gov/site/nypd/

stats/crime-statistics/crime-statistics-landing.page.

———. “Midtown Manhattan Security Initiative.” Press release, NYPD, September 20,
2010. http://www.nyc.gov/html/nypd/html/pr/pr_2010_midtown_security_
initiative.shtml.

———. Patrol Guide: Command Operations. 212–123. New York: NYPD, January 8,
2018. https://www1.nyc.gov/assets/nypd/downloads/pdf/public_information/
body-worn-cameras-patrol-guide.pdf.

———. Patrol Guide: “Padlock Law” Program. 291–31. New York: NYPD, July 1,
2014. https://www1.nyc.gov/assets/nypd/downloads/pdf/public_information/
public-pguide3.pdf.

New York State Division of Criminal Justice Services. Operation of License Plate
Readers in New York State: Suggested Guidelines. Albany: NYS DCJS, June
2017. https://www.aclu.org/files/FilesPDFs/ALPR/new-york/alprpra_
rensselaercountysheriffdepartment_troyny_2.pdf.

New York State Governor’s Office. “Governor Cuomo Announces More Than 100
Arrests since Major Enhancement to DMV’s Facial Recognition Technology.”
August 24, 2016. https://www.governor.ny.gov/news/governor-cuomo-
announces-more-100-arrests-major-enhancement-dmvs-facial-recognition-
technology.

New York State Senate. “Senate Bill S23.” Last modified December 28, 2016.
https://www.nysenate.gov/legislation/bills/2017/S23.

Nissenbaum, Helen Fay. Privacy in Context: Technology, Policy, and the Integrity of
Social Life. Palo Alto, CA: Stanford University Press, 2010.

Nolan, Lucas. “Facebook Could Face Billions in Fines over Facial Recognition
Features.” Breitbart. April 17, 2018. http://www.breitbart.com/tech/2018/04/
17/facebook-could-face-billions-in-fines-over-facial-recognition-features/.

NYPD News. “Finest Technology Helps New York’s Finest Take Two Guns off the
Streets.” December 8, 2015. http://nypdnews.com/2015/12/finest-technology-
helps-new-yorks-finest-take-two-gun-off-the-streets/.

95
O’Reilly, Kathleen. “Transparency, Accountability, and Engagement: A Recipe for
Building Trust in Policing.” Master’s thesis, Naval Postgraduate School, June
2017.

Orr, Steve. “New York Knows Where Your License Plate Goes.” USA Today, July 28,
2014. https://www.usatoday.com/story/news/nation-now/2014/07/28/new-york-
archiving-license-surveillance-data/13261679/.

People’s Daily Online. “‘Skynet’ System Supported by Facial Recognition Technology

Boosts Chinese Public Safety.” March 26, 2018. http://en.people.cn/n3/2018/
0326/c90000-9441798.html.

Perez, Bien. “Shanghai Subway to Use Alibaba Voice and Facial Recognition
Technologies.” South China Morning Post, December 5, 2017. http://www.scmp.
com/tech/enterprises/article/2123014/shanghai-subway-use-alibaba-voice-and-
facial-recognition-systems-ai.

Peterson, Andrew H., Jesse L. Kirkpatrick, and Deborah A. Boehm-Davis. Developing

Ethical, Legal, and Policy Analyses Relevant to the Use of Machine Learning
Algorithms in National Security. White paper, George Mason University, 2017.
http://sites.nationalacademies.org/cs/groups/dbassesite/documents/webpage/dbass
e_179882.pdf.

Phantom Plate. “Avoid Red Light and Speed Camera Tickets.” Accessed April 20, 2018.
https://www.phantomplate.com/.

Phillips, P. Jonathan, Patrick J. Flynn, Keven W. Bowyer, Richard W. Vorder Bruegge,

Patrick J. Grother, George W. Quinn, and Matthew Pruitt. “Distinguishing
Identical Twins by Face Recognition.” In Face and Gesture, 185–192.
Piscataway, NJ: IEEE, 2011. https://doi.org/10.1109/FG.2011.5771395.

Photo Blocker. “Avoid Red Light and Speed Camera Tickets.” Accessed April 21, 2018.
https://www.photoblocker.com/photoblocker.html.

Quintin, Cooper, and Dave Maass. “License Plate Readers Exposed! How Public Safety
Agencies Responded to Major Vulnerabilities in Vehicle Surveillance Tech.”
Electronic Frontier Foundation. October 28, 2015. https://www.eff.org/deeplinks/
2015/10/license-plate-readers-exposed-how-public-safety-agencies-responded-
massive.

Ramachandra, Raghavendra, and Christoph Busch. “Presentation Attack Detection

Methods for Face Recognition Systems: A Comprehensive Survey.” ACM
Computing Surveys 50, no. 1 (March 20, 2017): 1–37. https://doi.org/10.1145/
3038924.

96
Riffat, Muzamil. “Legal Aspects of Privacy and Security: A Case- Study of Apple versus
FBI Arguments.” Sans Institute. June 1, 2016. https://www.semanticscholar.org/
paper/Legal-Aspects-of-Privacy-and-Security%3A-A-Case-Study/
5f9139b8aa661bd040eec464d16ce31fbb829af5.

Risen, Tom. “Could the FBI See Your Selfies?” U.S. News & World Report, July 8, 2014.
https://www.usnews.com/news/articles/2014/07/08/fbi-may-seek-facebook-data-
for-facial-recognition.

Robbins, Liz. “Activists and ICE Face Off over Detained Immigrant Leader.” New York
Times, January 12, 2018. https://www.nytimes.com/2018/01/12/nyregion/
immigration-activist-deportation.html.

Roberts, David J., and Meghann Casanova. Automated License Plate Recognition Use by
Law Enforcement: Policy and Operational Guide, Summary. Doc. 239605.
Alexandria, VA: International Association of Chiefs of Police, September 2012.
https://www.ncjrs.gov/pdffiles1/nij/grants/239605.pdf.

San Diego County Sheriff’s Department. “Home Page.” Accessed May 29, 2018.
http://www.sdsheriff.net/tmh/docs/tmh-english.pdf

Sandoval, Greg. “Over 100 Amazon Employees, Including Senior Software Engineers,
Signed a Letter Asking Jeff Bezos to Stop Selling Facial-Recognition Software to
Police.” Business Insider, June 22, 2018. http://www.businessinsider.com/over-
100-amazon-employees-sign-letter-jeff-bezos-stop-selling-facial-recognition-
software-police-2018-6.

Savransky, Rebecca. “De Blasio: NYPD Planning to Have Body Camera on Every Cop
by Year’s End.” The Hill, January 31, 2018. http://thehill.com/homenews/state-
watch/371582-de-blasio-says-nyc-planning-to-have-every-cop-equipped-with-
body-camera.

Seaskate. The Evolution and Development of Police Technology. Washington, DC:

Seaskate, July 1, 1998. https://www.ncjrs.gov/pdffiles1/Digitization/
173179NCJRS.pdf.

Shay, Jim. “License Plate Reader Help Nab Suspect in Slaying of NYPD Explorer.”
Officer. July 11, 2018. https://www.officer.com/command-hq/technology/traffic/
lpr-license-plate-recognition/news/21012850/license-plate-reader-help-nab-
suspect-in-slaying-of-nypd-explorer-lesandro-junior-guzmanfeliz.

Siff, Andrew. “New MTA Towers Can Read License Plates, and Maybe More.” NBC
(New York). September 28, 2017. http://www.nbcnewyork.com/news/local/MTA-
Bridge-and-Tunnel-Gateway-Towers-Can-Read-License-Plates-Security--
448568193.html.

97
Silverman, Eli B. NYPD Battles Crime: Innovative Strategies in Policing. Boston:
Northeastern University Press, 1999.

Skirpan, Michael, and Tom Yeh. “Designing a Moral Compass for the Future of
Computer Vision Using Speculative Analysis.” In 2017 IEEE Conference on
Computer Vision and Pattern Recognition Workshops, 64–73. Piscataway, NJ:
IEEE, 2017. http://openaccess.thecvf.com/content_cvpr_2017_workshops/w16/
papers/Yeh_Designing_a_Moral_CVPR_2017_paper.pdf.

Stanley, Jay. Police Body-Mounted Cameras: With Right Policies in Place, a Win for All.
Version 2.0. New York: American Civil Liberties Union, March 2015.
https://www.aclu.org/other/police-body-mounted-cameras-right-policies-place-
win-all.

Statista. “Number of Motor Vehicles Registered in the United States from 1990 to 2016
(in 1,000s).” Accessed June 3, 2018. https://www.statista.com/statistics/183505/
number-of-vehicles-in-the-united-states-since-1990/.

Sunflex Zone. “SunflexZone Anti-ALPR & Red-Light Camera Privacy Solutions.”

Accessed June 9, 2018. https://www.sunflexzone.com/.

Sydell, Laura. “It Ain’t Me, Babe: Researchers Find Flaws in Police Facial Recognition
Technology.” NPR. October 25, 2016. http://www.npr.org/sections/
alltechconsidered/2016/10/25/499176469/it-aint-me-babe-researchers-find-flaws-
in-police-facial-recognition.

Tempey, Nathan. “The NYPD Is Tracking Drivers across the Country Using License
Plate Readers.” Gothamist. January 26, 2016. http://gothamist.com/2016/01/26/
license_plate_readers_nypd.php.

Townsend, Mark. “Police to Use Facial-Recognition Cameras at Cenotaph Service.”

Guardian, November 11, 2017. http://www.theguardian.com/technology/2017/
nov/12/metropolitan-police-to-use-facial-recognition-technology-remembrance-
sunday-cenotaph.

UK Parliament. “Protection of Freedoms Bill.” Accessed August 8, 2008.

https://publications.parliament.uk/pa/cm201011/cmpublic/protection/memo/pf11.
htm.

U.S. Congress. House of Representatives. Hearing before the Committee to Review Law
Enforcement’s Policies on Facial Recognition Technology. 115th Cong., 1st sess.,
March 22, 2017. https://oversight.house.gov/hearing/law-enforcements-use-facial-
recognition-technology.

98
U.S. Immigration and Customs Enforcement. Acquisition and Use of License Plate
Reader Data from a Commercial Service. DHS-ICE-PIA-039. Washington, DC:
Department of Homeland Security, December 2017. https://www.dhs.gov/
publication/dhs-ice-pia-039-acquisition-and-use-license-plate-reader-data-
commercial-service.

Vocativ. “Memo: New York Called for Face Recognition Cameras at Bridges, Tunnels.”
Last modified January 27, 2017. http://www.vocativ.com/396745/memo-new-
york-called-for-face-recognition-cameras-at-bridges-tunnels/.

Voelz, Glenn J. Rise of iWar: Identity, Information, and the Individualization of Modern
Warfare. New York: Skyhorse Publishing, 2018. https://ssi.armywarcollege.edu/
pubs/display.cfm?pubID=1292.

Waddell, Kaveh. “How License-Plate Readers Have Helped Police and Lenders Target
the Poor.” Atlantic, April 22, 2016. https://www.theatlantic.com/technology/
archive/2016/04/how-license-plate-readers-have-helped-police-and-lenders-
target-the-poor/479436/.

Waitt, Tammy. “CUBIC Testing Face Recognition for Subway (Learn More, Video).”
American Security Today. October 10, 2017. https://americansecuritytoday.com/
cubic-testing-face-recognition-subway-learn-video/.

Wexler, Chuck. “Introduction.” In How are Innovations in Technology Transforming

Policing, i–ii. Washington, DC: Police Executive Research Forum, January 2012.
http://www.policeforum.org/assets/docs/Critical_Issues_Series/how%20are%20in
novations%20in%20technology%20transforming%20policing%202012.pdf.

White, Michael D., and Henry F. Fradella. Stop and Frisk: The Use and Abuse of a
Controversial Policing Tactic. New York University Press, 2016.

Wood, Matt. “Some Quick Thoughts on the Public Discussion Regarding Facial
Recognition and Amazon Rekognition This Past Week.” AWS New Blog, June 1,
2018. https://aws.amazon.com/blogs/aws/some-quick-thoughts-on-the-public-
discussion-regarding-facial-recognition-and-amazon-rekognition-this-past-week/.

Woodward, John D., Jr., Christopher Horn, Julius Gatune, and Aryn Thomas. Biometrics:
A Look at Facial Recognition. Santa Monica, CA: RAND Corporation, 2003.
http://www.dtic.mil/dtic/tr/fulltext/u2/a414520.pdf.

99
THIS PAGE INTENTIONALLY LEFT BLANK

100
 INITIAL DISTRIBUTION LIST

1. Defense Technical Information Center

Ft. Belvoir, Virginia

2. Dudley Knox Library

Naval Postgraduate School
Monterey, California

101