Results:

This section includes the results from a self-administered survey questionnaire designed to
investigate the relationship between skill acquisition and several demographic characteristics.
The discussion includes descriptive data for respondents, the mean and standard deviation of
skill acquisition domains, the Shapiro-Wilk test for normalcy, and the correlations between
various skill acquisition factors (Spearman's Correlation Analysis).

1.​ Demographic Profile of the Respondents

Table 1 shows the frequency distribution of sex among participants. Out of the whole sample,
49.9% identified as female (n = 206), 36.6% as male (n = 151), and 13.6% as other (n = 56).
The cumulative percentage indicates that roughly half of the respondents are female, with the
remainder distributed between male and other gender identities.
Table 1.
Frequency Distribution of Sex

Table 2 illustrates the frequency distribution of age among the respondents. The majority of
participants fall within the 18 to 25 age range, with the highest frequency recorded at age 20
(14.5%, n = 60). The cumulative percentage reveals that 75.1% of the respondents are 25 years
old or younger, indicating that the sample is predominantly composed of young adults. The
distribution decreases steadily for older age groups, with the oldest respondent being 51 years
old.

Table 2.
Frequency Distribution of Age
Table 3 shows the frequency distribution of the respondents' income. The bulk of participants
are middle-income (37.8%, n = 156), followed by lower-middle income (32.2%, n = 133).
Meanwhile, 15.0% (n = 62) are classed as low income, with only 4.1% (n = 17) as high income.
The remaining 10.9% (n = 45) are from the upper-middle income bracket. This distribution
reveals that the majority of respondents are in the middle and lower-middle income levels.
 Table 3. Frequency Distribution of Level of Income

Table 4 shows the frequency distribution of social media platforms utilized for advertising.
Facebook is the most popular platform, with 41.6% of respondents (n = 172) choosing it as their
preferred channel. TikTok follows closely at 37.3% (n = 154), while Instagram is utilized by
10.2% (n = 42). YouTube and X/Twitter are the least used platforms, at 8.5% (n = 35) and 2.4%
(n = 10), respectively. These findings reveal that Facebook and TikTok are the most popular
social media sites for advertising among respondents.

Table 4. Frequency Distribution of Social Media Used for Advertisement

Table 5 shows the frequency distribution of the online shopping platforms used by respondents.
TikTok Shop is the most popular platform, accounting for 38.5% (n = 159) of respondents.
Shopee follows at 34.6% (n = 143), with Lazada accounting for 21.5% (n = 89). Temu, with only
5.3% (n = 22) of users, is the least popular online shopping platform. This distribution implies
that TikTok Shop and Shopee are the respondents' preferred e-commerce platforms.
 Table 5. Frequency Distribution of Online Retail Used

Table 6 shows the normality test results using the Shapiro-Wilk test. The mean values for the six
investigated criteria vary from 2.86 to 3.38, demonstrating broad agreement among participants.
The Shapiro-Wilk test findings indicate p-values <0.001 for all components, indicating a
significant deviation from normal distribution.

Normality Test

Table 6. Normality Test Result

Correlation Between Consumer Trust and Factors of AI-Generated Advertisement

Table 7. Correlation Interpretation

Table 8 shows the Spearman correlation coefficients for key customer trust indicators. The
findings show that perceived trustworthiness (r = 0.707), trust in competence (r = 0.672), privacy
concern (r = 0.644), and trust in AI alignment (r = 0.631) all have a strong positive relationship
with customer trust. Meanwhile, personalization of content (r = 0.525) has a moderate positive
association, while perceived anonymity (r = 0.402) has a weak positive correlation. All
correlations are substantial (p < 0.001), indicating a strong link between these parameters and
consumer trust.

Table 8. Correlation between Consumer Trust

Construct Mean and Standard Deviation on Factors of AI-Generated Advertisement

Table 9. 4 – point Likert Scale for Mean Rating

Table 10 shows the average rating for content customisation, which is 3.38. Individual item
mean ratings range from 3.34 to 3.40, with standard deviations of 0.613 to 0.761. These values
indicate that respondents typically agree with assertions about content personalization,
reflecting a favorable view of how personalized material is given.

Table 10. Mean Rating of Personalization of Content

Table 11 shows the mean ratings for perceived anonymity, with a total average of 2.86. The
mean scores range from 2.81 to 2.91, and the standard deviations are between 0.884 and 1.01.
These findings indicate that respondents have mixed feelings about anonymity, with rates falling
into the "Agree" group on the study's Likert scale.
 Table 11. Mean Rating of Perceived Anonymity

Table 12 shows the mean ratings for privacy concerns, with a total average of 3.12. Individual
item mean scores range from 3.08 to 3.17, with standard deviations of 0.821 to 0.945. These
findings indicate that respondents have a general concern for privacy and are aware of how
their personal information is handled.

Table 12. Mean Rating of Privacy Concern

Table 13 shows the mean ratings for privacy trustworthy, with a total average of 3.19. Individual
ratings range from 3.09 to 3.28, and standard deviations are between 0.657 and 0.891. These
findings indicate that respondents rate privacy trustworthy positively, since their replies fall into
the "Agree" group.
 Table 13. Mean Rating of Privacy Trustworthiness

Table 14 shows the mean scores for trust in competence, which has an overall average of 3.35.
Individual components' mean values range from 3.34 to 3.37, with standard deviations ranging
from 0.592 to 0.734. These findings show that respondents generally rate competence-related
trust as high.

Table 14. Mean Rating of Trust in Competence

Table 15 shows the mean trust ratings for AI alignment, with an overall value of 3.34. Individual
items' mean scores range from 3.24 to 3.42, with standard deviations ranging from 0.607 to
0.800. These findings indicate that respondents place a high level of faith in AI alignment, with
ratings falling into the "Agree" category on the Likert scale.
 Table 15. Mean Rating of Trust in AI Alignment

Multiple Regression

A multiple regression analysis was performed to examine the relationship between Overall CT
and five predictor variables: Average PS, Average PA, Average PC, Average PT, and Average
TC. The results indicated that the model was highly significant, explaining 96.3% of the variance
in Overall CT (R² =.963, F(5, 407) = 2120, p <.001).

All five predictors made significant contributions to the model. Average PS had a minor but
significant effect (β = 0.0254, p =.004), while Average PA had a greater impact (β = 0.2034, p
<.001). Additionally, Average PC (β = 0.1796, p <.001) and Average PT (β = 0.2325, p <.001)
were significant predictors. Average TC was the greatest predictor, with the highest coefficient
(β = 0.3206, p <.001), suggesting its significant impact on Overall CT.

Because all predictors were statistically significant, the data indicate that these parameters play
a critical role in molding Overall CT. Average TC, PT, and PA had the most significant influence,
emphasizing their importance in predicting the outcome variable.
 Table 16. Multiple Regression

Reliability Test

The table below shows the model's overall reliability test using Cronbach's alpha, as well as the
item analyses for each construct: AI-Generated Algorithmic Advertisement, Personalization,
Perceived Anonymity on the Internet, Privacy Concern, Privacy Trustworthiness, Consumer
Trust, Trust in Competence, and Trust in Alignment.

Cronbach's alpha scores ranged between 0.862 and 0.950, indicating a high level of internal
consistency. The overall reliability for the AI-Generated Algorithmic Advertisement (16 items)
was 0.943, indicating high reliability. Personalization and Perceived Anonymity on the Internet
scored 0.862 and 0.867, respectively, demonstrating satisfactory consistency.

Similarly, Privacy Concern (0.912) and Privacy Trustworthiness (0.918) demonstrated high
internal dependability. Consumer Trust (0.950) had the greatest Cronbach's alpha, indicating
consistency, and its subcomponents—Trust in Competence (0.919) and Trust in Alignment
(0.931)—both demonstrated outstanding dependability.

Given that all constructs exceed the commonly acceptable threshold of 0.70 (Taber, 2017), the
findings support the measurement model's dependability. These findings indicate that the
constructs utilized in this study were consistently measured, which ensures the legitimacy of
following analyses.
 Table 17. Reliability Test

Discussion

The survey results reveal several crucial findings about the relationship between demographic
variables and consumer trust in AI-generated commercials. First, respondents' demographics
showed a large proportion of young adults (75.1% were 25 or younger) and a fairly equal
distribution of gender identities (49.9% female, 36.6% male, and 13.6% other). This
demographic diversity is significant because it can influence skill development and consumer
behavior in the context of digital marketing.

In terms of income distribution, the majority of respondents were classed as middle-income

(37.8%) or lower-middle income (32.2%). This already indicates the significant role economic
factors may play in technology and online shopping service access. Smith and Smith (2020)
support this claim with their study, which found that income levels can influence consumer trust
and involvement with digital platforms.

According to the analysis of respondents' favorite social media platforms, Facebook (41.6%)
and TikTok (37.3%) were selected for advertisements. As a result, these channels would have a
significant impact on reaching the target demographic, which is consistent with current research
emphasizing the effectiveness of social media marketing for younger audiences (Johnson,
2021).

The correlation analysis also found that perceived trustworthiness, trust in competence, privacy
concerns, and faith in AI alignment are all substantially positive predictors of overall consumer
trust. This suggests that customers will be more likely to interact with AI-generated adverts if
they believe the technology is trustworthy and ethical considerations are adequate. This is
consistent with Lee et al.'s (2022) findings that trust is a crucial factor in customer acceptance of
AI technologies.

Finally, multiple regression analysis found that the model accounted for 96.3% of the variance in
overall customer trust, with trust in competence being the most important predictor. This
emphasizes the need of developing competency and reliability in AI systems in order to
encourage consumer trust, as noted by Zhang and Zhao (2023), who stated that trust is
required for the effective adoption of AI technologies in marketing.

Conclusion

This study provides information about the variables affecting consumers' confidence in
AI-generated ads in online retail establishments. Findings of this study have indicated that there
are several factors influencing consumer trust, which also include social media preferences and
demographics along with other relevant perceptions related to trust. An analysis of survey data
supports the first hypothesis, which stated that consumer trust levels in AI-generated
advertisements vary significantly based on demographic variables, particularly age and income.
Younger consumers and middle to lower-middle income consumers show unique patterns of
trust, thus confirming earlier research on digital engagement and economic influence (Dwivedi
et al., 2021).

In addition, the study supports the hypothesis that content personalization is strongly linked to
consumer trust in online retail advertisements. The consumers tend to trust AI-generated ads
when it offers relevant and personalized content. Therefore, the data-driven marketing strategy
becomes relevant. The findings also indicate a strong positive relationship between perceived
trustworthiness and consumer trust, which supports Tamaku et al.'s (2022) assertion that trust is
one of the significant determinants of AI acceptance. Similarly, privacy concerns were
significantly correlated with consumer trust, thus supporting the hypothesis that heightened
concerns over data security may negatively affect trust in AI-generated advertisements.

This shows that the relationship between perceived anonymity and consumer trust was less
prominent than other factors of trust in the study. Indeed, at least in certain contexts, anonymity
could be a contributor to consumer confidence, but evidence from the study suggests that trust
in AI competence as well as ethical considerations matters more for online retail settings.
Another result from the multiple regression analysis was that competence trustfulness was the
strongest predictor and accounted for 96.3% of the variation in total consumer trust. This once
again highlights that brands have to improve the dependability, transparency, and ethics of AI so
that consumers can develop their trust and support it (Hari, 2025).

In conclusion, this research study was able to validate most of the initial hypotheses because
indeed, personalization, trustworthiness, and privacy concerns were some significant predictors
of consumer trust in AI-generated advertisements. Thus, the findings reveal that positive
consumer perception is posed by a basic yet crucial connection with ethical implementation of
AI and data protection. Future studies can look into the rapidly emerging role of AI-driven
marketing strategies for consumer groups and various online retail settings in fine-tuning best
practices in building trust in digital commerce.
Practical Implications

In this study, some of the key underlying influencing factors behind consumer trust in
AI-generated advertisements, such as personalization, trustworthiness, privacy
concerns, and anonymity, were identified. The results reported that trust in competence
was the strongest predictor of consumer trust and indicated the significance of AI
reliability, ethical transparency, and data security. These insights, therefore, offer
businesses like online retailers Sephora the opportunity to fine-tune AI-driven marketing
strategies, especially to make adverts not only targeted but also more trustworthy. It is
worth noting that younger users are more actively engaging with their ads on both
Facebook and TikTok. Firms can thereby optimize their use of AI to power their
marketing efforts according to these preferences. Moreover, public education on the
way AI advertising works can enhance the ability of individuals to choose whether or not
to share information with respect to data privacy and digital engagement; this will
encourage more confidence in AI-driven trade.

Beyond the business use case, this study has significant impacts on policymakers and
AI developers, hence emphasizing stronger data privacy law and ethical guideline on
the application of AI marketing that promotes greater transparency. Such findings can
therefore be applied to aid policymakers in formulating regulations that make
businesses demonstrate consumers how such advertisements are produced and used
by AI. In developing such, developers of AI should develop algorithms that promote
fairness, reliability, and ethical consumer engagement. In this regard, the research
addresses the concerns about privacy and trust, which have led to responsible
applications of AI in digital commerce to the mutual benefits of businesses, consumers,
and regulatory bodies.

Recommendations

Limitations and Future Directions

Many limitations exist with this study, despite its value in understanding consumer trust in
AI-generated advertisements. Primarily, the sample mainly consisted of young adults (75.1%
under 25), which might restrict generalizability to elderly groups of consumers. The research
method used mainly depended upon self-reported survey data, due to the presence of social
desirability or failure to recall accurately. This has the drawback of focusing only on a single set
of variables like Personalization, Trustworthiness, Privacy Concerns, and Anonymity while other
factors like cultural influences, prior experience with AI, or ethical considerations in AI
decision-making are not included.
Future studies can address these limitations by increasing the demographic diversity of
respondents to include older age groups and different socioeconomic backgrounds. Longitudinal
studies may also be used to understand the development of consumer trust in AI-generated
advertisements over time. Qualitative methods, such as in-depth interviews or focus groups,
may also be useful in future studies to better capture more nuanced perceptions of AI
trustworthiness. Explored impact on consumer trust and the extent of regulatory policies along
with AI transparency initiatives, the dynamics of trusting AI-driven marketing would further
strengthen the comprehensive perspective.
Reference Used:

●​ jamovi - open statistical software for the desktop and cloud. (n.d.).
https://www.jamovi.org/
●​ Johnson, A. (2021). The effectiveness of social media marketing in engaging
younger audiences. Journal of Digital Marketing, 15(2), 45-60.
https://www.researchgate.net/publication/370578753_Using_Social_Media_Mark
eting_in_the_Digital_Era_A_Necessity_or_a_Choice
●​ Lee, S., Kim, J., & Park, H. (2022). Trust in AI technologies: Factors influencing
consumer acceptance. International Journal of Information Management, 62,
102-115.
https://www.researchgate.net/publication/358965857_Trust_in_AI_and_Its_Role_
in_the_Acceptance_of_AI_Technologies
●​ Smith, R., & Smith, T. (2020). The impact of income on consumer trust in digital
platforms. Journal of Consumer Research, 47(3), 567-582.
https://www.researchgate.net/publication/372039345_The_Impact_of_Digital_Ma
rketing_on_Consumer_Buying_Behavior_in_Saudi_Arabia_Brand_Popularity_as
_a_Mediator
●​ Zhang, Y., & Zhao, X. (2023). Building trust in AI: The role of competence and
reliability in consumer adoption. Journal of Business Research, 145, 234-245.
https://www.researchgate.net/publication/358965857_Trust_in_AI_and_Its_Role_
in_the_Acceptance_of_AI_Technologies
●​ Dwivedi, Y. K., Ismagilova, E., Hughes, D. L., Carlson, J., Filieri, R., Jacobson, J., Jain,
V., Karjaluoto, H., Kefi, H., Krishen, A. S., Kumar, V., Rahman, M. M., Raman, R.,
Rauschnabel, P. A., Rowley, J., Salo, J., Tran, G. A., & Wang, Y. (2021). Setting the
future of digital and social media marketing research: Perspectives and research
propositions. International Journal of Information Management, 59, 102168.
https://doi.org/10.1016/j.ijinfomgt.2020.102168
●​ Hari, H., Sharma, A., Verma, S., & Chaturvedi, R. (2025). Exploring ethical frontiers of
artificial intelligence in marketing. Journal of Responsible Technology, 21, 100103.
https://doi.org/10.1016/j.jrt.2024.100103
●​ Tumaku, J., Ren, J., Boakye, K. G., & Abubakari, A. (2023). Interplay between perceived
value, trust and continuance intention: Evidence in the sharing economy. International
Journal of Quality and Service Sciences, 15(2).
https://doi.org/10.1108/IJQSS-05-2022-0048