This article was downloaded by: [Eindhoven Technical University]

On: 07 July 2014, At: 15:16

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number:
1072954 Registered office: Mortimer House, 37-41 Mortimer Street,
London W1T 3JH, UK

Intelligence and National

Security
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/fint20

What's wrong with the

Intelligence Cycle
Arthur S. Hulnick
Published online: 22 Dec 2006.

To cite this article: Arthur S. Hulnick (2006) What's wrong with the
Intelligence Cycle, Intelligence and National Security, 21:6, 959-979, DOI:
10.1080/02684520601046291

To link to this article: http://dx.doi.org/10.1080/02684520601046291

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all
the information (the “Content”) contained in the publications on our
platform. However, Taylor & Francis, our agents, and our licensors
make no representations or warranties whatsoever as to the accuracy,
completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of
the authors, and are not the views of or endorsed by Taylor & Francis.
The accuracy of the Content should not be relied upon and should be
independently verified with primary sources of information. Taylor and
Francis shall not be liable for any losses, actions, claims, proceedings,
demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in
relation to or arising out of the use of the Content.
 This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,
reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access
and use can be found at http://www.tandfonline.com/page/terms-and-
conditions
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014
 What’s Wrong with the Intelligence Cycle

ARTHUR S. HULNICK

In the modern era, almost all intelligence professionals will study the
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

Intelligence Cycle as a kind of gospel of how intelligence functions. Yet

it is not a particularly good model, since the cyclical pattern does not
describe what really happens. Policy officials rarely give collection
guidance. Collection and analysis, which are supposed to work in
tandem, in fact work more properly in parallel. Finally, the idea that
decision makers wait for the delivery of intelligence before making
policy decisions is equally incorrect. In the modern era, policy officials
seem to want intelligence to support policy rather than to inform it. The
Intelligence Cycle also fails to consider either counter-intelligence or
covert action. Taken as a whole, the cycle concept is a flawed model,
but nevertheless continues to be taught in the US and around the world.

No concept is more deeply enshrined in the literature than that of the

‘Intelligence Cycle’. I studied the Intelligence Cycle as an undergraduate in
Sherman Kent’s book on strategic intelligence, and then later when I attended
the US Air Force Intelligence School in 1957.1 In 1965, in the Central
Intelligence Agency’s (CIA) training courses, I studied it yet again. When it
came time to start writing about intelligence, a practice I began in my later
years in the CIA, I realized that there were serious problems with the
Intelligence Cycle.2 It is really not a very good description of the ways in
which the intelligence process works. Secondly, it ignores two main parts
of intelligence work, counter-intelligence and covert action. There is an
alternative view.
The Intelligence Cycle is so nicely described elsewhere3 that there seems no
need to go over it here. So, let us start at the beginning and look at what is
wrong. The notion that policy makers or intelligence consumers, as they are
sometimes called, provide guidance to intelligence managers to begin the
intelligence process is incorrect. Policy consumers do sometimes indicate
their main concerns to intelligence managers, but often they assume that the
intelligence system will alert them to problems, or provide judgments about

Intelligence and National Security, Vol.21, No.6, December 2006, pp.959 – 979
ISSN 0268-4527 print 1743-9019 online
DOI: 10.1080/02684520601046291 ª 2006 Taylor & Francis
 960 INTELLIGENCE AND NATIONAL SECURITY

the future. Consumers will sometimes tell intelligence managers what they are
worried about, or the direction in which they intend to take policy, but not
always.
Still, it is usually not too difficult for intelligence managers to learn what
policy makers are up to, though the managers often have to take the initiative
to obtain the information. If intelligence managers at various levels are in
touch with their policy counterparts, this sharing of information may work
quite well. Over the years, intelligence managers have tried to systematize
this process by asking policy officials to provide specifics on their concerns.
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

In the Carter administration, for example, a system of National Intelligence

Topics (NITs) was created as a way of soliciting guidance for intelligence.
Later they were called Key Intelligence Questions (KIQs). In some cases,
when policy consumers failed to submit NITs or KIQs, managers had to
resort to sending policy officials a list of topics, asking them to cross out the
ones they thought were not necessary, or adding those they wanted to the list.
Even then the lists were sometimes ignored.
In the end, intelligence managers have to make decisions about the subjects
that ought to be covered. Often this is driven by world events. But none of
this provides guidance for intelligence collection. The guidance comes from
within the system. Secretary of Defense Donald Rumsfeld, in the George W.
Bush administration, is once reported to have said that ‘we don’t know what
we don’t know’, but that is usually not the case. Intelligence managers often
know what gaps exist in the intelligence data base derived from intelligence
collectors and analysts. Filling the gaps is what drives the intelligence
collection process, not guidance from policy makers. Thus, the first step in
the intelligence cycle is incorrect in reality.

THE SECOND STEP

The second step is equally incorrect. Collection managers cannot wait for
guidance in regard to gaps in the intelligence data base to begin the collection
process. The gaps will be filled once the collection process is underway. For
example, in running espionage operations, commonly called HUMINT (for
HUman INTelligence), it may take months or years to find a person who has
access to the information needed, and is willing to be recruited as a spy. The
same may be true for technical collection sensors. Satellites in space, which
make up many of the sensor platforms, are not nearly as flexible as managers
would wish. Thus, anticipating the intended targets cannot be overlooked. For
example, during the British confrontation with Argentina over the Falkland
Islands, the US could not help the British with space imagery because the
satellite, programmed to observe the Soviet Union at that time, only passed
over the Falklands at night.
 WHAT’S WRONG WITH THE INTELLIGENCE CYCLE 961

Of course, with the use of unmanned aerial vehicles, imagery collection

has become more easily re-focused on targets of opportunity, but the
unmanned aircraft may still not be in the right place when they are needed.
Even open source intelligence (OSINT), which has been given new life in
recent years because of the proliferation of information on the Internet,
requires planning to insure access to needed material. Intelligence managers
need sophisticated software to mine the data because there is so much of it.

THE REAL DRIVERS

Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

For all these reasons, intelligence managers, and not policy officials, are the
real drivers of the intelligence collection process. Clearly, intelligence moves
from collection to analysis, as the Intelligence Cycle holds, but analysts do
not always need new intelligence material to understand world events. The
data base is already so large that a competent analyst could write about
most events without any more than open sources to spur the process. The
incremental addition of new intelligence from human sources or technical
sensors may modify the analytic process but rarely drives it.
The job of the analyst is, in part, to evaluate raw material and put it in
perspective. The analyst receives intelligence material from a variety of
sources, including media reports, official reports from other government
agencies, as well as reports from the intelligence collection process. In my
experience as a practicing analyst in the military and in CIA, raw reports
from human sources or technical sensors are sometimes fragmentary, biased,
contradictory, or just plain wrong. In order to analyze the data, the analyst
compares the new material with the existing data base and previous analysis.
Hanging a finished product, whether it is current reporting or a longer range
estimate, on one source usually does not work well. This is apparently what
happened in the case of the estimate on weapons of mass destruction (WMD)
in Iraq that helped trigger the invasion of that country. The estimate was
based, in part, on the reporting of one rather poor and unreliable source. The
estimate turned out to be quite wrong, as we now know.4

OPERATING IN PARALLEL

A better way of looking at the relationship of intelligence collection and

intelligence analysis is to think of the two processes as operating in parallel
rather than sequentially. The two processes are co-equal in terms of utility. It
is important to note as well that raw reporting from the collection process, set
up into standardized formats, usually goes to policy officials as well as to
analysts at about the same time. While this may not always be true in other
intelligence systems, it is certainly true in the US. While senior policy
 962 INTELLIGENCE AND NATIONAL SECURITY

officials may not see a great deal of the raw reporting, there are usually watch
centers at the various policy agencies that screen the raw reporting and send
forward the most interesting ones.
Unfortunately, as I have already noted, some of this raw intelligence may
be incomplete, contradictory, or just wrong. Policy officials sometimes take
the reporting as having been judged and evaluated. Thus, I have heard
officials say that CIA has reported an event when in fact what the officials
have seen is an unevaluated agent report passed along to them by their watch
centers. It is not possible to stop this flow of raw reporting. As Bob Gates, the
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

former Director of CIA once noted to me, once the spigot is opened it is not
possible to close it, even though allowing consumers to have raw reporting at
about the same time as the analysts receive them creates some serious
problems for the analysts. Collection managers often take a different view.
They believe they are doing a great service to the policy community by
providing this raw reporting. If the Intelligence Cycle really worked, the
circulation of raw reports to policy officials would not happen.

A MAJOR PROBLEM

Since intelligence collection and intelligence analysis operate in parallel and

should be co-equal, one would expect that there would be a great deal of
information sharing between the two. Regrettably, this is not always the case.
Because of restrictions of information sharing, psychological barriers, fears
of compromising sources, and security concerns, the intelligence collection
process and the intelligence analytic process not only operate in parallel, they
are sometimes quite independent of each other. This is a major problem.
When I first joined the CIA, I was assigned on a temporary basis to an
office in the Directorate of Plans (DDP), later renamed the Directorate of
Operations (DO), and in 2005, renamed yet again as the National Clandestine
Service (NCS). My job was to deal with incoming reports from the field.
When an interesting report came in one day, I asked my boss if we should
alert the relevant analyst about it. He rejected the idea, saying that our job
was to send reports like it to the White House, and not to the Directorate of
Intelligence, since analysts were not worth the attention. I was shocked.
Later, when I became an analyst, I did my best to establish good relations
with my operational colleagues, but there were issues.

BARRIERS TO COMMUNICATION

In those days, there were physical barriers, manned by armed guards, to

prevent analysts and operations officers from visiting each other’s offices.
Later, the physical barriers were removed, but the psychological ones
 WHAT’S WRONG WITH THE INTELLIGENCE CYCLE 963

remained. Ops people feared that, somehow, analysts would mishandle

reports from the field and reveal the identity of clandestine sources. Analysts
mistrusted operations officers because they were thought to be devious and
untrustworthy. This mistrust was kindled in part because analysts in those
days tended to be introverts who found the extroverted personality of the
typical operations officers to be abrasive. Ops people tended to think that the
introverted analysts were ‘wimps’.
Over the years these stereotypes have largely been overcome, but recent
efforts to increase communication between analysts and operators by co-
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

locating them have not always been successful. Agency managers have
pushed analysts to take tours overseas with field stations, but it is more
difficult for an operations officer to serve a tour as an analyst. Similar issues
may not arise in other intelligence agencies unless they have co-equal
collection and analysis components.
The Defense Intelligence Agency (DIA), for example, was immune to this
sort of ‘stovepipe’ problem because it was mostly an analytic rather than a
collection agency. As defense HUMINT grows, perhaps the same problem
will arise. The National Security Agency (NSA) and the National Geo-Spatial
Intelligence Agencies (NGIA) are devoted mostly to collection, but they
tend to be tightly compartmented, creating a different kind of ‘stovepipe’
problem. In theory, all the intelligence agencies should share raw data and
coordinate analysis, but for a variety of reasons they do not always do so. This
was one of the main critiques of both the 9/11 Commission and the
Commission investigating the intelligence failure surrounding the failed
estimate on WMD.

A MIXED BAG

There is a tendency among intelligence agencies to hold back the most

sensitive and exciting reports until the agency’s leaders have been able to
deliver the reports to senior policy officials, thus highlighting the skill and
cleverness of their people and scoring ‘points’ with the officials. One effort to
spur inter-agency communications has been the establishment of centers
where all the agencies have representation and where their representatives can
easily talk with their counterparts, even informally, to discuss events and
incoming intelligence. The establishment of these centers has been something
of a mixed bag. We know from the 9/11 investigations that the then existing
counter-terrorism center (CTC) was not a place where all information was
shared.5
Now efforts at intelligence reform have ‘morphed’ the CTC into the
National Counter Terrorism Center (NCTC), controlled by the new
Director of National Intelligence (DNI). Will that spur the agencies into
 964 INTELLIGENCE AND NATIONAL SECURITY

more easily sharing their best and most sensitive data? It would be nice to
think so, but experience shows that this does not always happen, even at the
highest level.

THE FINAL STAGES

In the final stages of the Intelligence Cycle, finished intelligence, broken

down into a variety of products, emerges from the analytic process, is
supposed to be delivered to policy officials – the literature refers to this
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

delivery as dissemination – and then policy officials either make decisions or

create further requirements and the cycle starts over again. This, too, is a
distortion of what really happens. Much of this depends on the kind of
intelligence product that is being delivered. These products include warn-
ing intelligence, in which consumers are alerted to ‘breaking news’, current
intelligence to update consumers on world events on which they already have
some knowledge, in-depth studies on particular situations or issues, and
forecasts of the future, the estimate. Each product is received and used in a
different way, but none of them really drive the policy process.
Warning intelligence is supposed to alert policy officials to ‘breaking’
world situations, especially those for which they may have to take action.
Both intelligence managers and policy consumers hate surprise. It is
embarrassing for intelligence when the system misses an event about which
it should have had information. For example, the CIA failed to detect the fact
that the Indian government planned to conduct a nuclear test in 1998. Later
investigations revealed that this was both a collection and analysis failure.
The CIA had no assets it could tap in India at that time, and the Indian analyst
at CIA had somehow missed the fact that the Indian prime minister had
declared his intention to hold the tests. Despite the fact that there was little
the US government could do to stop the tests, policy officials were
nonetheless outraged at this failure.6
Even greater outrage was directed at the entire Intelligence Community for
its failure to detect the 9/11 terrorist attack on the US. This has been
exhaustively examined and has led to the restructuring of the intelligence
system. Yet there is considerable evidence that there was little that might
have been done to avert the disaster. However, it illustrates the point that
policy officials expect the intelligence system to be all-knowing, all-seeing,
and always correct. As Richard Betts pointed out many years ago,
intelligence failure is probably inevitable.7
Warning of crisis should come early enough so that policy officials can
have time to develop some kind of considered response. Unfortunately, the
warning may come so late that it is really an alert that the crisis has already
begun. Using a system that is comprised of warning centers at major military
 WHAT’S WRONG WITH THE INTELLIGENCE CYCLE 965

commands, tied in to warning centers at all the intelligence agencies and in

policy departments in Washington, and taking advantage of the proliferation
of 24-hour television and Internet outlets, the warning network rarely misses
the start of a crisis, and is then able to reach out to decision makers quite
rapidly. When the decision makers ask intelligence officers how they should
respond to the crisis, typically intelligence officers decline to provide advice,
thus staying clear of the policy process.

THE MOST USEFUL PRODUCT

Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

Current or daily intelligence is the most ubiquitous of all types of intelligence

products, delivered at all levels and usually first thing in the morning. It is
designed to supplement the media, based on the assumption that policy
officials have already got their media inputs from newspapers or television
news. It is the most popular of all intelligence products because current
intelligence is an ‘easy read’, short, and to the point. For those policy officials
who only have 10 or 15 minutes a day to absorb intelligence products – and
consumer surveys consistently show that this is about all the time policy
officials have for such things – current intelligence is rated as the most useful
product from the Intelligence Community. The idea of this product is to
summarize events, explain how they fit into some context, and suggest what
might happen next. It is a very journalistic methodology.
Unlike warning intelligence which may lead to policy action, as the
Intelligence Cycle suggests, current intelligence hardly ever leads to policy
decisions, and is not meant to do so. Instead, it gives generalists at senior
levels a chance to find about events outside their main areas of responsibility.
Specialists often complain that the daily intelligence flow does not provide
the level of detail they would need to make policy, but the current intelli-
gence products are not designed for specialists. In fact, it would be quite
likely that specialists would have seen a great deal of the raw intelligence
data that lay behind the current intelligence product anyway.
During the 9/11 investigation, much was made of the fact that one daily
publication, the President’s Daily Brief (PDB), had on 6 August 2001 reported
the possibility that terrorists might use commercial aircraft as cruise missiles
to attack commercial or government buildings within the US.8 Critics of the
president took this to have been a warning the president and his senior
staff had missed, but normally the PDB would not have been the kind of
intelligence product used for warning. The warning would have been
delivered in a much more specific document devoted entirely to the subject.
Intelligence managers have never expected the PDB or similar publications
to be more than educational in nature. Certainly, they do not drive the
intelligence process.
 966 INTELLIGENCE AND NATIONAL SECURITY

IN-DEPTH STUDIES

The same might be said for the myriad in-depth intelligence studies churned
out by the analytic components. These studies have proliferated in recent
years, although they were rarely attempted at the beginning of the Cold War.
These studies are designed to provide in-depth analysis on specific subjects
and are meant more for policy officials at working levels rather than senior
decision makers, who rarely have the time to read them. These studies are
helpful in forcing analysts to come to grips with a specific subject, provide
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

useful information to consumers within the intelligence system, and are

supportive of policy makers as they design policy initiatives.
The production of these studies grew over the years as a way of giving
analysts a vehicle for attacking a problem in more depth than was possible in
a daily or weekly publication, and without the fuss and bureaucracy involved
in producing the more formal national estimate. Policy officials sometimes
request these in-depth studies, along the lines suggested by some versions of
the Intelligence Cycle, but in many cases the studies are produced because
analysts are directed by intelligence managers to write them, or analysts
themselves believe they should be written. When Robert M. Gates took over
the Directorate of Intelligence (DI) at CIA during the early days of the
Reagan administration, he decreed that analysts should produce at least two
of these in-depth studies every year. Gates was fond of pointing out that the
DI produced about 5,000 of these studies one year. It was not clear, however,
how many of them were actually read.
In recent years, these studies have been more carefully tailored to the needs
of policy officials. The same might be said for the Defense Intelligence
Agency products, which are geared to military needs, or those coming from
the State Department’s intelligence and research unit, which has always
focused its analysis on foreign policy issues. The fourth category of product,
the estimate, is the one most likely to drive the policy process, at least in
theory, but the reality is often different.

THE REALITY OF ESTIMATES

The estimate is a creature of the Cold War, but has its roots in World War II.
It is supposed to be a forecast of the future which decision makers can use to
build policy, just as the Intelligence Cycle proposes. The estimate is supposed
to be drawn by analysts from all the producing agencies, coordinated by the
analysts among themselves to reach an agreed forecast, with dissenting views
included. Then it is blessed by the agency leaders, signed off at the top and
sent to the president, the National Security Council and staffs, and serves
as the basis for policy discussions. There are actually cases where this has
 WHAT’S WRONG WITH THE INTELLIGENCE CYCLE 967

happened, where decision makers have waited for the Intelligence

Community’s views as embodied in the estimate, but these cases are rare.
The reality is that policy officials often know what they want to do even
before they receive the estimate, and hope that this product will confirm in
some way the wisdom of the path they have already chosen. When the
estimate conflicts with their views, policy consumers may dismiss it as
uninformed, useless, or even obstructionist. When it agrees with what they
think they already know, then they may see it as confirming, irrelevant, or
again useless. Although one would think that policy makers would want to
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

know when they were heading in the wrong direction, this is not usually the
case. Policy consumers do not welcome intelligence that is non-confirming,
perhaps because the large egos that brought them into positions of power do
not permit admissions of ignorance.

THE WMD CASE

There is no better example of what can go wrong in the estimates process

than the recent experience with the problems related to Iraq and Saddam
Hussein’s alleged possession of weapons of mass destruction. We now know
how the intelligence system politicized the estimate to meet the needs of the
George W. Bush administration. The estimate on WMD was flawed from the
beginning. It was based on the reporting of only a few unreliable sources.
Then analysts made several faulty assumptions about the weapons Saddam
Hussein had had or used before the first Gulf War. Finally, policy officials
used the estimate to convince both Americans and other nations that Saddam
was about to develop nuclear weapons. All of this was wrong.9
According to James Risen, intelligence officials in both the collection and
analysis arms of CIA, as well as those in other agencies, knew the sources
were poor and the conclusions wrong, but they could not fight senior
managers who wanted to satisfy the political needs of the White House. Even
more corrupting, it appears that Secretary of Defense Rumsfeld, fearing that
the estimate would not support the already planned invasion of Iraq, sent
his own officials, neither of them intelligence officers, to find the ‘correct’
information.10 All these steps were perversions of the estimates process. One
can only hope that such antics will not take place in the future.
For all the reasons I have cited, it seems clear to me that trying to learn
how intelligence works by using the Intelligence Cycle model will lead to
misunderstandings about what really happens in the intelligence world.
Collection and analysis are actually parallel processes. The key to their
effective functioning lies in the extent to which there is good communication
between the two processes. There needs to be, as well, good communication
between intelligence managers and policy consumers throughout the
 968 INTELLIGENCE AND NATIONAL SECURITY

intelligence process. At the same time, however, intelligence managers

must stand up to policy officials when they seek to make the intelligence
judgments conform to political needs.

INTELLIGENCE AND POLICY

In the early days of the Cold War, the founders of the CIA debated the extent
to which intelligence should be close to policy. Sherman Kent, a Yale
professor who went on to establish the national estimates system in CIA, and
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

one of the early thinkers about the intelligence process, believed, as did ‘Wild
Bill’ Donovan and others, that if intelligence became enmeshed in the policy
process it would lose its value. Kent argued that the best way to avoid
politicization of intelligence was to remain distant and aloof. Later, Roger
Hilsman, one of the intelligence chiefs at the State Department, took a
different view. Hilsman thought that intelligence had to be close to policy to
remain relevant.11 The experiences of the Bush era suggest that Kent may
have been right all along.
Nonetheless, other studies have shown that there must be good communi-
cation between policy consumers and intelligence managers if intelligence is
to be on target and meet the needs of decision makers. At the same time,
intelligence managers have to stand up to efforts by policy officials to skew
intelligence judgments when the conclusions are at variance with the political
proclivities of partisan officials. No one said this would be easy. It is a con-
stant challenge to provide ‘truth to power’. Intelligence must deliver the
unvarnished bottom line. Policy officials can go elsewhere for politicized
information if they wish, but at their own peril.

LOOKING AT COUNTER-INTELLIGENCE

Leaving aside the collection and analysis processes in intelligence, one

cannot understand the entire intelligence system without looking at counter-
intelligence. Counter-intelligence is largely defensive in nature, and is not
part of the traditional intelligence cycle, although some writers have tried to
adapt the cycle into a counter-intelligence model.12 In my view, counter-
intelligence follows an entirely different and unique path, with a model of its
own. It is certainly worth studying because counter-intelligence is a major
function of intelligence, consists of both active and passive components, and
has become as controversial as any aspect of the intelligence function in
government.
In its earliest forms, counter-intelligence usually meant counter-espionage,
stopping enemy, adversary, or even friendly spies from stealing a country’s
own secrets. Of course, the target country might very well be carrying out
 WHAT’S WRONG WITH THE INTELLIGENCE CYCLE 969

espionage against the enemies, adversaries, or friends at the same time as it

tries to defend against similar sorts of spying. Thus, stealing secrets for one’s
country is good and necessary; having one’s secret’s stolen is dangerous and
despicable. US intelligence officers, for example, are rewarded for their
successes in gathering information from their targets, even though some of
what they do may be illegal in the countries they target. At the same time,
other US intelligence officers are heralded for their ability to root out foreign
spies and are castigated when they fail to do so. After all, espionage is illegal
in the US and must be stopped.
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

MORE DIVERSE

Today, counter-intelligence has become much more diverse than just

stopping spies. It now means countering terrorism, narcotics flows, global
organized crime, and subversion. Whatever the threat, the patterns of intelli-
gence activity in fighting all of them are similar. It has nothing to do with
the intelligence cycle. Instead, there is a counter-intelligence methodology
that is unique.
First, in countering national security threats, counter-intelligence units
must identify and locate the evil-doers. These might be foreign intelligence
operatives working for a hostile intelligence service, a terrorist cell, a unit of
a crime ‘army’, or a group of narcotics pushers. There are several proven
intelligence methods for identifying the ‘bad guys’, including the use of
penetrations, or ‘moles’, to get on the inside of the groups or services,
surveillance, either physical or technical, the use of informants, and
intelligence derived from captured or detained individuals. All have both
positive and negative aspects.
Based on the Cold War experience, we know that it is possible to recruit
officials of a foreign intelligence service to become turncoats and betray
some of the activities of their operatives. There are several known cases
where the US was able to place a mole inside a foreign service, and there
were a number of US intelligence officers, such as Aldrich Ames, John
Walker, and Robert Hanssen, who gave away US secrets to the Soviets. The
Federal Bureau of Investigation (FBI) seemed quite capable of recruiting
penetrations of crime groups such as the Sicilian Mafia. Penetrating a terrorist
cell is far more difficult and dangerous. Terrorist cells are usually made up of
a handful of people, all of whom may be bonded by family or religious ties.
Even if a terrorist cell member wanted to become a ‘double agent’, the first
hint of disloyalty to the cell could result in death.
Physical or electronic surveillance is another proven method of identi-
fying counter-intelligence targets. Overseas, this kind of surveillance can be
mounted against potential targets as a result of decisions by intelligence
 970 INTELLIGENCE AND NATIONAL SECURITY

managers. In the US, however, the rules are stricter. Counter-intelligence

officials would, in most circumstances, be required to go through a legal
process and obtain a warrant before employing surveillance against a US
citizen, a resident alien, or a US person. This issue became front page news
early in 2006 when the New York Times revealed that President George W.
Bush had authorized surveillance of communications without warrant, argu-
ing that the Congress had given the president the authority to do so.13 The
issue may not be resolved until a court case is brought, or new legislation is
passed defining the parameters of surveillance use domestically.
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

USING INFORMANTS

Informants can be very useful in identifying counter-intelligence targets.

Informants are not recruited agents, but rather people who see something
amiss and report their suspicions to authorities. In hostage situations,
informants may be able to point out where unusual activity is taking place.
For example, prior to 9/11, flight school managers reported to the FBI their
concerns about Middle Eastern men seeking flight training only to steer
aircraft, rather than learn to take off and land. Unfortunately, FBI senior
officials refused to grant field agents permission to interview the informants,
claiming that there was no probable cause to do so.14
Informants can also cause a lot of wasted effort. During the sniper crisis in
Washington DC in 2002, where two men were able to terrorize the area by
random attacks on innocent targets, requests for information resulted in more
than 100,000 inputs, of which 40,000 were worth investigating.15 People who
have experience in fielding informant reports note that often the reports are
used to denounce spouses, parents, or unpleasant neighbors, and provide no
useful intelligence. Nonetheless, informants can prove to be helpful in
identifying bad guys.

INTELLIGENCE FROM INTERROGATION

After 9/11, a good deal of controversy arose over the use of intelligence
gained from the interrogation of detainees, either overseas or here at home.
Some men of Middle Eastern Muslim extraction who were not US citizens
were required to register with the federal government. Some of these people
had irregularities in their visas, had overstayed in the US, or were in the US
illegally. They were detained in somewhat harsh conditions and, in some
cases, badly mistreated. It did not appear, however, that much effort was
made to find out if any of them had ties to terrorism.
At the same time, as the US geared up to take down the Taliban
government in Afghanistan, some Taliban fighters or people associated with
 WHAT’S WRONG WITH THE INTELLIGENCE CYCLE 971

Al Qaeda were captured, turned in by informants, or sold to the US by

Afghan warlords. These people, dubbed ‘enemy combatants’ by the Bush
administration, were shipped to the Guantanamo Naval Base in Cuba, where
US authorities said US legal rules did not apply to them. These people were
interrogated using what some described as harsh methods, or even torture,
according to press reports.
After the US-led invaded Iraq in 2003, more detainees were captured on
the battlefield. These fighters were imprisoned in Iraq at some of Saddam
Hussein’s former prisons, including the infamous one at Abu Ghraib. It was
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

at this location that the worst abuses took place. Apparently, unschooled,
unscreened, untrained guards were turned loose to abuse the prisoners in the
mistaken belief that this would ‘soften them up’ for interrogation. All of these
situations involving detainees were handled badly.
Long experience has taught that there are effective ways to interrogate
prisoners, using methods that do no harm to the subjects while produ-
cing useful intelligence. Unfortunately, those lessons were not applied
effectively in the post-9/11 situations. The literature on interrogation
methods, on training interrogators, on handling subjects should have been
readily available to anyone involved in trying to extract intelligence from
detainees. Anyone who has been involved in intelligence-style interrogations
knows that torture is ineffective and counter-productive, as well as abhorrent
and illegal. Since those experiences, the rules have been changed to exclude
such behavior by US officials.
A good interrogation may yield only bits and pieces of information, but if
intelligence collectors are careful, they may be able to piece together a
broader picture from a series of subjects. The main aim, of course, is to try to
learn something about the cells, or units that the subjects have come from,
especially about their plans for future operations.

STOPPING THE BAD GUYS

After the ‘bad guys’ have been identified, a decision has to be made about the
kinds of operations that will be mounted to stop whatever kind of plan or
activity might be underway against US interests. This creates a dilemma.
Usually, intelligence officers will press to extend or broaden the collection
effort to make sure that all the ‘bad guys’ have been identified and located. At
the same time, law enforcement officials are eager to bring the ‘bad guys’ to
justice. This creates a serious problem, often described as the ‘cops and spies’
dilemma.16
The divisions between law enforcement and intelligence in the US have
deep roots. Unlike many other industrialized countries, the US does not have
a domestic intelligence service, such as the MI-5 in Great Britain, or the
 972 INTELLIGENCE AND NATIONAL SECURITY

DST in France. Instead, the US has relied for many years on the FBI, which is
really a law enforcement organization, to gather counter-intelligence and
then act to bring lawbreakers to justice. In other countries, the domestic
intelligence services collect and analyze counter-intelligence in parallel with
counterpart foreign intelligence organizations, which work beyond the
country’s borders. When suspected criminal behavior is uncovered, the
domestic intelligence services may turn to national police organizations to
carry out law enforcement operations against the suspects.
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

INTELLIGENCE VERSUS LAW ENFORCEMENT

In the US, however, where no domestic intelligence service has existed, there
have been both legal and procedural barriers between the national intelli-
gence services, whose focus has been almost exclusively abroad, and the FBI,
which has always had a role in domestic counter-intelligence. Traditionally,
counter-intelligence collected abroad was passed to the FBI, which then
determined, usually in consultation with the Justice Department, whether or
not there was probable cause to open a criminal investigation. This would be
used to gather evidence that could be brought if a court case arose. This was
different from the gathering and analysis of intelligence data, which
traditionally was not treated or handled as evidence.
Because of cover considerations and the need to protect the identity of
intelligence officers, intelligence managers did not want their people to have
to appear in court, and wanted as well to protect the sources and methods
used to collect intelligence. The FBI was under no such strictures, but the
evidence gathered had to be backed by appropriate warrants and protected
according to legal standards. The Aldrich Ames case is a perfect example of
how this system used to work.
A joint CIA and FBI team was able to track down Ames and identify him
as a Soviet mole in the CIA. Then the FBI obtained a warrant under the
Foreign Intelligence Surveillance Act (FISA) to be certain that they had the
right target. Once that was done, a second FBI team, with a criminal warrant,
took over the investigation and gathered evidence that might be used to
prosecute Ames. In the end, Ames agreed to a plea bargain and the case never
came to court, but it illustrates how a firewall was in place to separate the
counter-intelligence investigation from the criminal one.17

TAKING DOWN BARRIERS

Since 9/11, some the barriers between intelligence and law enforcement have
been weakened, but the cops and spies dilemma still exists. Under the new
rules, the FBI may levy requirements on the US intelligence services to
 WHAT’S WRONG WITH THE INTELLIGENCE CYCLE 973

collect information specific to their domestic needs. It is not yet clear if such
intelligence would be used as evidence in court cases. More likely, the
FBI would ask that intelligence be gathered to support its newly created
National Security Service, which combines the FBI’s older counter-terrorism
and counter-intelligence units with its newer intelligence bureau, created after
9/11.
At the same time, the CIA and the FBI have drawn more closely together
with a strong push from Congress. FBI agents have been assigned to CIA for
counter-intelligence purposes for many years, and more recently to fight
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

terrorism. CIA officers are reportedly working closely with FBI field offices
where antiterrorism task forces have been created. While this may break
down traditional barriers between the two agencies, there is still some
resentment among CIA officers about the growing role of FBI attaches
serving abroad, and FBI concerns about the reluctance of CIA officers to
share information.

THE COUNTER-INTELLIGENCE MODEL

So, when one looks at the pattern of counter-intelligence functions, it

does not look at all like the Intelligence Cycle. Instead, it may be seen as
follows:

IDENTIFICATION
PENETRATION
EXPLOITATION
INTERDICTION
CLAIM SUCCESS

In this pattern, exploitation is the process of learning as much as possible

about the ‘bad guys’ before moving against them. Interdiction means either
arresting the law breakers or pre-empting their operations. While political
leaders often talk about bringing the enemy to justice, suggesting that they
would be arrested and taken to trial, convicted, and punished, in the fight
against terrorism, pre-emption may be the preferred course of action, espe-
cially overseas. In one case, for example, a US Predator with a missile on
board was reportedly used to strike a terrorist leader in Yemen, killing him
and his associates while they were driving in the desert.18 One might argue
that this was punishment before trial, or, alternatively, that this was necessary
to prevent the terrorist from leading a strike against the US.
The downside of pre-emption is that sometimes innocent victims are slain
along with the intended targets. That was apparently what happened when the
Bush administration launched a missile from a drone aircraft against Ayman
 974 INTELLIGENCE AND NATIONAL SECURITY

al-Zawahiri, Osama bin Laden’s deputy, early in 2006. The missile killed
17 people, according to press reports, but not the intended target, who later
broadcast an attack on Mr Bush, equating him with Adolf Hitler. Despite the
failure to kill Zawahiri, the Bush administration later said that the attack had
indeed killed an important terrorist leader and was worth the cost.19

EXPLOITATION BEFORE INTERDICTION

In the Counter-Intelligence Model, exploitation comes before interdiction,

Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

meaning that as much intelligence should be gathered before the case or

operation is turned over to law enforcement. Of course, in cases where there
is pressure to stop the enemy or adversary, exploitation may come before the
intelligence is fully gathered. For example, in the case of the so-called
‘Lackawanna Six’, exploitation was cut short because of the need for poli-
tical leaders to show that they were cracking down on terrorism. The Six
were Yemeni immigrants living outside Buffalo, New York, who went to
Afghanistan before 9/11 in the misguided belief that training with the Taliban
fighters was going to be something of a lark.20
When they discovered that the Taliban were really training terrorists, the
Yemeni immigrants returned to the US. After 9/11, they turned themselves in
to authorities to explain what they had done. They were quickly arrested, and
eventually jailed. No one at the time seemed to realize that at least one or
two of them might have been sent back to Afghanistan as double agents to
penetrate Al Qaeda. Even an effort to learn more about their experiences
was cut short by the pressure to achieve quick convictions to show that the
government was moving swiftly against terrorism.

CLAIMING SUCCESS

Finally, in the last step of the counter-intelligence process, authorities often

make public claims of success, a rare step in intelligence work. Normally,
intelligence managers try very hard to keep successes secret so that they
might be repeated. An oft quoted CIA saying is ‘The secret of our success is
the secret of our success’.21 In cases where intelligence has been gathered
successfully, it is critical to protect sources and methods. In counter-
intelligence, however, the claim of success, made when the case has ended,
could be used to convince the public that the government is ever watchful and
actually doing something with the billions of dollars spent on intelligence.
During his tenure as FBI Director, J. Edgar Hoover made a fine art out of
going public with counter-intelligence success. His senior agents all received
training in public relations and the FBI was made to look good, even when
serious mistakes had been made.22
 WHAT’S WRONG WITH THE INTELLIGENCE CYCLE 975

While intelligence is usually carefully hidden, except for the counter-

intelligence cases, intelligence failure quickly becomes public. This is a
serious problem for intelligence managers. In the early days of the CIA, there
was no public affairs function even to deal with the public or the media.
When Admiral Turner became director, however, he instituted a Public
Affairs Office, much to the chagrin of many old-timers. Since then, the CIA
has had to wrestle with the appropriate response when media queries arise.
This is especially true when a spy case, such as the capture of Aldrich Ames,
becomes public knowledge, or when a covert action surfaces.
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

More forthcoming CIA directors, such as George Tenet, have had the
Public Affairs Office respond generously to media questions. Under the
successor regime of Porter Goss as CIA director, however, the CIA seemed to
return to a more conservative approach. In such cases, it would not be
uncommon for the media to receive the standard answer to questions about
intelligence. This says that the CIA ‘can neither confirm nor deny allegations
of intelligence activity’ which is little more forthcoming than ‘no comment’.
Nonetheless, enterprising reporters, such as Bill Gertz of the Washington
Times and James Risen of the New York Times seem to be quite successful in
learning about inside stories at CIA and other intelligence agencies.

DEFENSIVE COUNTER-INTELLIGENCE

There are defensive measures in counter-intelligence that do not fit either into

the traditional Intelligence Cycle or in the model I have just described. These
measures are often lumped together as various aspects of security. They
include careful background checks on prospective employees, including the
use of polygraph interviews to verify the information candidates submit on
their applications, and continuing monitoring of employees throughout their
careers. Facilities used for intelligence and other governmental functions are
extensively guarded and patrolled, monitored with alarm and surveillance
devices, and protected by barrier entry devices to keep out unwanted visitors.
Some facilities have protective systems in roadways and parking areas
that can be activated to stop suicidal vehicle bombers. Buildings may be
shielded electronically to prevent an adversary’s use of listening devices or
electronic surveillance to intercept and steal secrets. Most importantly,
employees are trained in security awareness, so that they can report anything
that seems to be a threat. They are taught to protect the secrets with which
they have been entrusted, and this responsibility lasts even after they leave
their employment.
For example, those of us who were once inside the system and signed
secrecy agreements are obligated to submit their published materials, inclu-
ding this one, to their agencies for review before they are given to their
 976 INTELLIGENCE AND NATIONAL SECURITY

editors and publishers. This is not censorship, but rather a system to ensure
that no secret information is inadvertently released. Some CIA authors
have taken advantage of this system to include blacked-out passages in their
books, demonstrating that they really were prepared to release sensitive
information, but were stopped by the review process. This tends to sell more
books and can be a clever marketing ploy.

THE COVERT ACTION FUNCTION

Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

The last function of intelligence, and again one not included in the
Intelligence Cycle, is that of covert action, or special operations. This activity
is not really intelligence in its traditional role of gathering and analyzing
information, but rather the use of intelligence resources to carry out the
national security policy of the state using surreptitious methods. Intelligence
agencies around the globe carry out such operations because they have the
necessary secret facilities and personnel. All through the Cold War, it was
covert action that drew most of the attention and most of the criticism of
American intelligence.
General Jimmy Doolittle, one of the notable heroes of World War II, after
taking a hard look at intelligence in the immediate post-war period,
concluded that the US would have to be more clever, more tricky, and more
devious that the Communist adversaries in order to overcome their bid for
world domination. He stated that Americans would just have to accept this
‘repugnant’ policy.23 When the CIA became involved in trying to overthrow
governments in Guatemala, Iran, Indonesia, and Cuba, and was severely
criticized in some quarters for having done so, it became clear that there were
limits to what the American people were prepared to accept. Much has been
written about the nature and limits of covert action and there seems no need
to repeat that here.24 While covert action does not fit into the Intelligence
Cycle, there is a pattern to this function worth outlining. This pattern is
similar to other aspects of policy development and implementation, except
that covert action is supposed to be secret and to disguise the role of the US.

POLICY FORMULATION

The pattern of policy formulation looks a bit like the Intelligence Cycle, but
is in reality quite different. In the first step of the policy process, policy
officials within the national security bureaucracy recognize and identify a
problem with which they must deal. Theoretically, the identification of the
problem comes from intelligence, but in reality policy officials often see this
at about the same time as intelligence officials because both receive the
incoming data at about the same time, as I have earlier explained.
 WHAT’S WRONG WITH THE INTELLIGENCE CYCLE 977

In the next step, policy officials begin to seek options for dealing with the
problem, assuming some role for the US is necessary. At this point, one of the
options might well be a covert action. We know from long experience with
covert action that it only makes sense as an adjunct to policy and should
not be the policy itself.25 Thus, the choice of using covert action remains
with decision makers and is not chosen by intelligence. The conventional
wisdom in some circles during the Cold War was that intelligence managers
decided to mount covert actions independent of policy officials. This notion
that the CIA was a ‘rogue elephant’ running amok was debunked during the
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

famous investigations of intelligence held by Senator Frank Church in the

1970s. Church learned that all CIA covert actions had been directed in some
way by the White House and funded in secret by members of Congress.26
Finally, decision makers at the top choose the option they desire, and direct
its implementation. In the case of covert action, this requires that the
president issue a written finding that the covert action is needed. Then the
appropriate intelligence official must brief the Intelligence Oversight
Committees of the Congress, in secret, about the policy ‘on a timely basis’.
The Congress has often pressed presidents to issue the findings before the
option is implemented, but presidents have usually chosen to ignore this,
claiming that it infringes on their freedom of action.27
The pattern looks like this:

PROBLEM RECOGNITION
OPTION CREATION
OPTION SELECTION
IMPLEMENTATION

Intelligence analysis should feed into the process at all stages, but we know
that the options that policy officials choose are driven by many things, and
intelligence is not always at the top of the list.

CONSEQUENCES OF COVERT ACTION

Covert action has both short- and long-term consequences. This is true of all
kinds of policy choices, but because covert action is kept secret, the normal
debate about policy choices takes place among a relatively small group of
people. The result is that short-term solutions, which may seem attractive at
the time they are chosen, may prove to have unintended consequences in the
long run. There are too many examples to cover here, and the literature on
covert action is voluminous. Suffice to say that US governments rarely think
about the long-term consequences of policy choices and in that regard covert
action is no different than more open kinds of policies.
 978 INTELLIGENCE AND NATIONAL SECURITY

There is a long tradition in intelligence that intelligence officers do not

offer policy recommendations to decision makers. While this may be true
for the delivery of finished intelligence products, it is not so in regard to
covert action. As Dr James Steiner, a former CIA officer, has pointed out, in
covert action, especially in the war on terrorism, the attempt to be policy
neutral does not apply.28 For many years, a senior CIA officer has been
assigned to the White House Staff to help work out the details of covert
action when policy makers decide they need to have such operations. This
officer’s role is to make sure that requested covert actions are feasible and
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

supportable. Thus, the officer is as much a policy maker as an intelligence

official.29
I have argued elsewhere that one way to address the short-term versus the
long-term consequences of covert action is to set up a center, much like the
other inter-agency centers in US intelligence.30 This center, however, should
include both intelligence and policy officials. Its goal would be to analyze
how a covert action might work, and what its impact would be. The
intelligence officers assigned to such a center should come from both the
analytic and the operational units of the CIA. Traditionally, covert action has
been kept compartmented within operations units, without the benefit of
analytic inputs.
It seems pretty clear that presidents will always want to have the
option of using some form of covert action against enemies and adver-
saries. No presidents in living memory, even those who were suspicious of
covert action, have ever said that they would not use it. Therefore, the
intelligence agencies which might be involved in such operations, primarily
the CIA in the present US Intelligence Community, must be prepared
to be tasked to carry out covert action, and must maintain the capability to
do so.

A FLAWED VISION

I suspect that, despite my preaching about alternatives to the traditional

Intelligence Cycle, it will continue to be taught both inside government and
elsewhere. Nonetheless, it would be encouraging to think that those so deeply
wedded to the flawed concept of the Intelligence Cycle would, in the course
of studying this volume, realize that there is an alternative to the traditional
view of how intelligence works. Perhaps they might even consider it for
discussion. Yet we know that people tend to look for confirming rather than
disconfirming data. They will seek to defend the Intelligence Cycle rather
than consider the alternatives. Nonetheless, the Intelligence Cycle is a flawed
vision, and thus poor theory. One need only ask those who have toiled in the
fields of intelligence.
 WHAT’S WRONG WITH THE INTELLIGENCE CYCLE 979

NOTES
The author would like to thank the Praeger publishing house in the United States for allowing this
article to be published here as well as, simultaneously, in the work edited by Loch K. Johnson,
Strategic Intelligence, 5 vols. (Westport, CT: Praeger 2007).

1 Sherman Kent Strategic Intelligence for American World Policy (Princeton, NJ: Princeton
University Press 1966). I studied an earlier version published in 1948.
2 See, for example, Arthur S. Hulnick, ‘The Intelligence Producer-Policy Consumer Linkage:
A Theoretical Approach’, Intelligence and National Security 1/2 (May 1986) pp.212–33.
3 Johnson, Strategic Intelligence.
4 James Risen, State of War: The Secret History of the CIA and the Bush Administration (New
Downloaded by [Eindhoven Technical University] at 15:16 07 July 2014

York: Free Press 2006).

5 The 9/11 Commission Report (New York: W.W. Norton & Co. 2003) pp.339–60.
6 Arthur S. Hulnick, Fixing the Spy Machine: Preparing American Intelligence for the 21st
Century (Westport, CT: Praeger 1999) p.59.
7 Richard K. Betts, ‘Analysis, War, and Decision: Why Intelligence Failures Are Inevitable’,
World Politics 31 (1978).
8 Arthur S. Hulnick, Keeping Us Safe: Secret Intelligence and Homeland Security (Westport,
CT: Praeger 2004) p.16.
9 Risen, State of War.
10 Hulnick, Keeping Us Safe, pp.85–6.
11 Roger Hilsman, Strategic Intelligence and National Decision (Glencoe, IL: Free Press 1956).
12 ‘CSIS and the Security Intelligence Cycle’, 5http://www.csis-scrs.gc.ca4 (accessed 1 April
2004).
13 Risen, State of War, pp.39–60.
14 9/11 Report.
15 Arthur S. Hulnick, ‘Indications and Warning for Homeland Security: Seeking a
New Paradigm’, International Journal of Intelligence and CounterIntelligence 18/4 (Winter
2005–06) pp.593–608.
16 Hulnick, Keeping Us Safe, pp.103–18.
17 See, for example, Pete Earley, Confessions of a Spy: The Real Story of Aldrich Ames (New
York: G.P. Putnam’s Sons 1997).
18 Hulnick, Keeping Us Safe, p.72.
19 Craig Whitlock and Walter Pincus, ‘Qaeda Deputy Mocks Bush’, Washington Post, 31
January 2006.
20 Hulnick, Keeping Us Safe, pp.126–7.
21 Hulnick, Fixing the Spy Machine, p.81.
22 Ronald Kessler, The Bureau: The Secret History of the FBI (New York: St. Martin’s Press
2002).
23 Harold M. Greenberg, ‘The Doolittle Commission of 1954’, Intelligence and National
Security 20/4 (December 2005) pp.687–94.
24 See, for example, Abram Shulsky and Gary Schmitt, Silent Warfare (Washington, DC:
Brassey’s 2002).
25 James E. Steiner, ‘Restoring the Red Line Between Intelligence and Policy on Covert
Action’, International Journal of Intelligence and CounterIntelligence 19/1 (Spring 2006)
pp.156–65.
26 See, for example, Rhodri Jeffreys-Jones, The CIA and American Democracy, 3rd ed. (New
Haven: Yale University Press 2003).
27 Christopher Andrew, For the President’s Eyes Only (New York: Harper Collins 1995).
28 Steiner, ‘Restoring the Red Line’.
29 Hulnick, Fixing the Spy Machine, pp.82–3.
30 Ibid.