i An update to this article is included at the end

Computers in Human Behavior 165 (2025) 108549

Contents lists available at ScienceDirect

Computers in Human Behavior

journal homepage: [Link]/locate/comphumbeh

AI-driven personalization: Unraveling consumer perceptions in social

media engagement
Tanawat Teepapal
International College, Rajamangala University of Technology Krungthep, Bangkok, Thailand

A R T I C L E I N F O A B S T R A C T

Handling editor: Paul Kirschner This study advances our understanding of the impact of personalized stimuli driven by artificial intelligence on
consumer engagement in social media marketing. The research develops and examines an extensive S-O-R model,
Keywords: linking AI stimuli to customer perceptions of trust, privacy concerns, perceived usefulness, and, consequently,
Artificial intelligence consumer engagement. Structural equation modeling was utilized to examine the gathered data and evaluate the
Customer engagement
hypotheses. The results confirm the hypothesis that AI-enabled personalization positively influences trust, pri­
Personalization
vacy concerns, and perceived usefulness. Trust and perceived usefulness positively impact consumer engage­
Digital marketing
Consumer perception ment, while privacy concerns do not. Unexpectedly, AI-enabled personalization doesn’t significantly affect
customer engagement. By exploring the mediating roles of consumer perceptions, the results emphasize
perceived utility and trust as a significant mediating factor, underscoring its crucial contribution to fostering
positive interactions between users and technology. The research extends the SOR model in understanding AI’s
impact on consumer engagement, emphasizing trust and perceived usefulness as crucial mediators. For practical
implications, businesses in social media marketing should prioritize trust-building, enhance user experience,
address privacy concerns, and adopt a customer-centric approach. These insights provide valuable guidance for
navigating AI driven personalization dynamics in social media marketing.

1. Introduction such as, eBay, alibaba, amazon, rakuten, and [Link], along with
popular social media like twitter, facebook, and instagram, have driven
Rapid technological advancements have significantly changed the the social digitalization of retail businesses. Digital technology has
field of marketing over the past few decades. The rise of Web 2.0, known transformed marketing by reducing regulatory barriers, resulting in
for user-generated content, has led to a shift in marketing to a more significant cost savings, and enabling personalized promotional cam­
intricate and ever-changing landscape with the development of Web 3.0, paigns for individual clients (Bag et al., 2021).
focused on data-driven strategies (Erragcha & Romdhane, 2014). In the e-commerce industry, companies utilize artificial intelligence
Data-driven marketing has undergone a gradual evolution, leveraging (AI) platforms to extract valuable customer insights by analyzing pat­
technological advancements such as data analytics, machine learning, terns in customer behavior, data, and feedback (Krishnan et al., 2002).
web analytics, social media analytics, virtual reality, augmented reality, These platforms have become indispensable for monitoring shifting
and chatbots for customer support. These advancements have allowed consumer behaviors and preferences, often replacing or complementing
marketers to refine their strategies and customize their responses to traditional marketing methods. Customer relationship management
meet the needs of consumers. This evolution has facilitated personali­ (CRM), social interactions and media, digital content marketing, e-
zation, automation, data analysis, predictive analytics, and improved WOM, digital marketing analytics, and virtual marketing communities
consumer experiences and engagement (Bag et al., 2021; Krishen et al., are all important tools for meaningful consumer engagement
2021; Sakas et al., 2023). (Gołąb-Andrzejak, 2023).
Technological developments in retail have greatly influenced the Moreover, integrating AI into digital marketing has transformed
industry through the emergence of dynamic multichannel and omni- consumer-brand interactions and shopping experiences. AI enables
channel strategies (Jin & Shin, 2020). Major online retail platforms digital marketers to automate processes, create personalized

engagement: like, share,...

AI increases customer perceptions
This article is part of a special issue entitled: AI & marketing published in Computers in Human Behavior.
E-mail address: tanawat.t@[Link]. Customer perception --> Engagement

[Link]
Received 3 December 2023; Received in revised form 19 August 2024; Accepted 21 December 2024
Available online 26 December 2024
0747-5632/© 2024 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
T. Teepapal Computers in Human Behavior 165 (2025) 108549

advertisements, optimize campaigns, and track results, thereby insights into how technological and environmental factors influence
improving efficiency and effectiveness (Huang & Rust, 2020). AI in the human responses (Gao et al., 2023).
digital market has already been used by businesses to better meet The SOR model used in this study states that customer perceptions in
customer needs and preferences by personalizing emails, social media social network platforms serve as the organism, contributing to
postings, websites, and other marketing communication devices improved customer engagement through behavioral responses. In to­
(Haleem et al., 2022). Content personalization, through customized day’s digital age, the SOR model remains highly relevant. In artificial
content and targeted advertising, has emerged as a key driver of intelligence systems, data serves as stimuli, driving synthesis and
engagement, enabling more effective digital marketing strategies. enabling the system to learn and adapt, facilitating efficient and
Digital technology and AI greatly influence how consumers behave appropriate responses, whereas traditional computer systems rely on
and make purchasing decisions (Bag et al., 2022). Personalization is static, predefined instructions (programs) without adaptive learning
essential for connecting with, engaging, and co-creating with customers. (Perez-Vega et al., 2021).
Additionally, it is crucial in fostering positive attitudes and reinforcing Organizations are utilizing AI-enabled personalization as cognitive
interpersonal connections (Payne, Dahl, & Peltier, 2021a). Contempo­ stimuli, tailoring content to individual preferences, thereby enhancing
rary marketing segmentation strategies heavily rely on collecting and their marketing capabilities. This alignment with the SOR framework
analyzing preference and behavior data. This includes data from sources highlights how customers perceived of digital marketing act as the or­
like Facebook (e.g., likes, shares), Google searches, and mobile app ganism influencing online customer engagement behaviors. To provide
usage (Ma & Sun, 2020). a comprehensive understanding, we will now review relevant literature
Although digital marketing can improve user experiences and build on this topic.
consumer-brand connections, it also presents challenges regarding pri­
vacy issues in digital marketing approaches (Gao & Liu, 2022; Rog­ 2.2. Artificial intelligence (AI) driven personalization
geveen et al., 2021). Nevertheless, this study’s primary goal is to
investigate how AI driven personalization affects consumer perceptions In the field of online marketing, artificial intelligence (AI) is
and engagement within the context of digital marketing. Understanding increasingly recognized for its ability to drive personalized experiences
consumer attitudes is crucial as new technology like AI-driven person­ tailored to individual consumer preferences and behaviors. As com­
alization continues to evolve in digital marketing. This information is panies around the globe continue to advance in AI, machine learning,
essential for marketers adapting to the evolving digital environment. and deep learning technologies, the role of AI in comprehending and
There is a significant knowledge gap exists regarding how consumer responding to consumers is becoming increasingly vital. According to
perception affects AI-driven personalization and its connection to Kaplan and Haenlein (2019), artificial intelligence (AI) is characterized
customer engagement in digital marketing. (Ameen et al., 2021; Andr­ by its ability to comprehend and learn from outside data, as well as
zejak, 2023). This research fills this gap by examining consumers’ per­ adapt to meet specific objectives. Specifically, AI-driven personalization
ceptions of AI-driven personalization in digital marketing content across in digital marketing aims not only to identify opportunities and
various platforms and their level of engagement. Furthermore, it con­ comprehend customer behaviors but also to establish meaningful con­
tributes to strategic marketing research by bridging the divide between nections with consumers (Krishen et al., 2021). This form of personali­
AI-driven marketing practices and academic studies, improving our zation extends beyond traditional marketing approaches, offering
comprehension of how AI-enabled personalization impacts consumer superior strategies, models, processes, and services that influence
engagement. Therefore, our inquiry seeks to exploring three specific customer behaviors (Huang & Rust, 2020). Moreover, while the objec­
research questions. tive of AI in digital marketing is to enhance rather than replace human
judgment, its ability to customize products according to individual
RQ1. What impact does AI-enabled personalization in online market­
customer needs can significantly impact sales outcomes (Campbell et al.,
ing, specifically on social media platforms, have on the awareness of
2020).
customer perceptions (trust, privacy concerns, perceived usefulness)
Furthermore, AI’s role in facilitating engagement marketing by
and consumer engagement?
leveraging data from previous customer interactions enhances value
RQ2. How does AI-enabled personalization affect user engagement creation and influence across the customer journey (Payne et al.,
across social media platforms? 2021a). Specifically, in social media marketing, AI-driven personaliza­
tion enables the customization of content and precise targeting of
RQ3. Do customer perceptions of trust, privacy concerns, and
audience segments, drawing from a diverse range of data sources such as
perceived usefulness mediate the relationship between AI-enabled
search browsing patterns, purchase histories, and demographic infor­
personalization and customer engagement?
mation. However, it’s crucial to acknowledge the potential pitfalls of
personalization, including the risk of negative consumer reactions if it
2. Theoretical and literature review
encroaches on individual choices or privacy (Roggeveen et al., 2021).
2.1. Stimulus organism response (SOR) model
2.3. Consumer perceptions
The SOR model is a widely used model for examining how external
In this dynamic digital environment, trust, privacy concerns, and
stimuli impact an individual’s cognitive and emotional processes before
perceived usefulness are important factors influencing how consumers
influencing their behavior (Mehrabian & Russell, 1974; Tang and Zhang,
perceive and engage in online and social media marketing that uses AI to
2018). In this framework, the stimulus (S) acts as an antecedent, while
personalize content for them. The stimulus organism response (SOR)
the organism (O) functions as mediating processes bridging the gap
model states that a consumer’s emotional state affects their perception.
between the stimulus and behavioral responses (R) (Bag et al., 2022;
It is essential for businesses to properly understand the complex con­
Gao et al., 2023).
nections among these components regardless of whether they expect to
Within these mediating processes, it is essential to consider con­
successfully survive the rapidly changing landscape of digital marketing.
sumers’ emotional states, encompassing factors like pressure, feelings of
happiness or joy (pleasure), a sense of stimulation (arousal), and a
- Trust
perception of control or influence over the mediator (dominance) (Sung
et al., 2021). The SOR model has been utilized in different areas such as
Trust is a multifaceted and crucial concept that holds significant
marketing, service management, and information systems, offering
importance across various fields, particularly in e-commerce and

2
T. Teepapal Computers in Human Behavior 165 (2025) 108549

artificial intelligence (AI). In both domains, trust is a fundamental factor automatically see AI-driven personalization as valuable when they
in ensuring success. According to Ha and Stoel (2009) and Ameen et al. believe it to be useful and convenient. This perception of value serves as
(2021), trust can be defined as an attitude characterized by confident a key driver behind their acceptance and utilization of personalized
belief that potential vulnerabilities in a risky online situation will not be marketing strategies (Sadriwala & Sadriwala, 2022). Prior studies have
maliciously exploited. This definition emphasizes how trust influences suggested that consumer attitudes regarding the adoption of new tech­
how customers behave in online interactions and transactions. nology are significantly influenced by perceived usefulness (Na et al.,
Consumers’ confidence in online retailers and internet technology 2022).
plays an important role in how they make decisions, impacting their Moreover, there is empirical evidence to support the strong positive
perceptions of safety and security while making online purchases (Ha & correlation between marketing innovation and the perceived usefulness
Stoel, 2009). Uncertainty about the safety and security of digital plat­ of artificial intelligence (Maduku et al., 2023; Sadriwala & Sadriwala,
forms significantly contributes to customer unwillingness to make pur­ 2022). in the competitive e-commerce landscape, businesses need to
chases online (Gefen & Straub, 2003). Previous studies have shown that communicate the tangible benefits of AI-driven personalization to
users’ trust in the online platform significantly influences their confi­ attract customers and improve their digital marketing efforts. Conse­
dence in participation in social commerce (Li, 2019). quently, perceived usefulness emerges as an important factor influ­
Moreover, trust becomes even more complicated when it comes to encing consumer behavior within the realm of AI and digital marketing.
AI-driven personalization, fundamentally reshaping consumers’ per­ After examining the current literature, we propose the following
ceptions and interactions with content. It extends beyond trusting the hypothesis.
technology and the brand of the retailer; it also encompasses having
H1. AI-enabled personalization significantly impacts customer trust
confidence in the objectives and processes behind AI-generated content
perceptions.
(Huang et al., 2022). Consequently, we consider trust to be an essential
factor in understanding consumer perceptions of digital online market­ H2. AI-enabled personalization significantly affects customer percep­
ing and how it influences consumer involvement with AI-enabled tions regarding privacy concerns.
personalization.
H3. AI-enabled personalization significantly impacts customer per­
ceptions of perceived usefulness.
- Privacy concern

Leveraging customer personal information is a powerful tool for

2.4. Consumer engagement
companies, enabling them to create precise and targeted marketing
campaigns that strengthen relationships with and retain customers. This
Customer engagement extends into the digital context, particularly
data-driven approach enhances targeting and relevance, resulting in
on social networking websites within e-tailer platforms, where users
increased engagement and brand affinity (Maduku et al., 2023). Benefits
interact with businesses and express their preferences through online
from customer relationships are offset, however, by growing concerns
comments and reviews (Kannan, 2017). This engagement is measured
about confidentiality and information security.
through consumer interactions such as likes, shares, and comments on
Gutierrez et al. (2019) define privacy as the ability to control and
posts, providing valuable insights through data analytics tools. Being
manage access to different areas of one’s social and private spaces, such
engaged in social media is crucial for companies as it enables them to
as information privacy, interpersonal relationships, physical bound­
access a larger customer base. LinkedIn, Facebook, Instagram (IG), and
aries, and psychological states. Individual attitudes and beliefs about
Twitter are popular social media platforms for evaluating customer
privacy can vary depending on the situation and have significant con­
engagement with online business campaigns (Bag et al., 2022). Ac­
sequences on how people share their data. Consumer perceptions of
cording to Bag et al. (2022), consumer engagement refers to methods,
privacy play a pivotal role in determining their willingness to disclose
processes, and technological applications that enable ongoing interac­
information when interacting with websites or online platforms. Nega­
tion with prospective clients across various channels.
tive perspectives can arise from increased concerns regarding data
Digital marketing tools play a significant role in enabling businesses
collection, storage, and possible exploitation of personal data (Alzaidi &
to engage with online customers (Pantano & Vannucci, 2019). However,
Agag, 2022). As data collection of customer expands, businesses of all
effectively engaging consumers through social media requires a deep
sizes face challenges from the growing number of security breaches. The
insight of target markets, compelling content, and a thoughtful approach
privacy issues significantly affect how willing customers are to use social
(So et al., 2021). Additionally, in today’s digital marketing context, AI
media platforms for shopping and sharing information. (Maduku et al.,
systems have become essential tools for businesses. However, their ef­
2023; Swani et al., 2021). However, despite privacy concerns, earlier
ficacy depends on consumer acceptance (Kaplan & Haenlein, 2019).
studied have shown that customers are still willing to embrace person­
This necessitates consumers acknowledging and agreeing with AI sug­
alized online marketing campaigns in order to enjoy their advantages.
gestions before fully embracing the system. With AI’s capacity to gather
(Swani et al., 2021).
and analyze data, it assists businesses in tailoring marketing campaigns,
Therefore, AI-enabled personalization in interactive marketing holds
fundamentally altering consumer interactions with organizations. The
the potential for advantages, but it also carries the risk of negative re­
transformation in customer engagement has a profound impact on
actions if it constrains choices or violates individuals’ privacy. There­
stakeholders’ psychological states throughout the entire process of
fore, we consider privacy and security to be crucial elements in
developing, producing, and consuming products or services, ultimately
comprehending how consumers engage in online marketing and inves­
influencing the co-creation of value (Gao et al., 2023).
tigating their influence on consumer behavior in AI-driven
Customer engagement involves a wide range of activities, including
personalization.
interactions on social media, active participation in loyalty programs,
and sustained interest in a company’s products or services. This
- Perceived Usefulness
increased level of engagement not only enhances a company’s financial
performance but also increases customer satisfaction, brand loyalty, and
According to Davis (1989), perceived usefulness refers to consumers’
growth and innovation opportunities (Kumar & Kalse, 2021). Therefore,
belief that technology will enhance their effectiveness, performance, or
hypotheses are proposed.
productivity in job-related tasks. This study examines how consumers
view the relevance and value of AI-driven personalized content that is H4. Customer trust perceptions significantly impact customer
customized to their specific needs (Gao & Liu, 2022). Customers engagement.

3
T. Teepapal Computers in Human Behavior 165 (2025) 108549

H5. Customer privacy concern perceptions significantly influence published studies: AI-enabled personalization (Lalicic & Weismayer,
customer engagement. 2021), consumer perception of trust (Sharma et al., 2021), perception of
privacy concerns (Gao et al., 2018), perceived usefulness (Ashfaq et al.,
H6. Perceived usefulness significantly impacts customer engagement.
2020), and consumer engagement (Hollebeek et al., 2014). A rating
H7. AI-enabled personalization significantly affects customer scale, with response options ranging where ’1′ indicating totally disagree
engagement. and a ’5′ indicates totally agree, was adopted to assess all constructs.

2.5. Mediating mechanisms 3.3. Validity and reliability

Van der Heijden et al. (2003) investigate the impact of perceived risk To assess the validity of each construct, we conducted a sampling
and perceived ease of use on attitudes on online purchasing attitude. The adequacy test and sphericity test in exploratory factor analysis. The
researchers identified these factors as crucial in understanding in­ KMO value resulted 0.916, which was higher than the suggested
dividuals’ intentions to make online purchases. In our study’s concep­ threshold of 0.9, and sphericity test’s significant result (p < 0.001),
tual framework, we build upon this insight and propose that AI-enabled indicating that the sampling was appropriate for factor analysis. Prin­
personalization’s impact on consumer engagement is mediated through cipal component analysis with varimax orthogonal rotation method was
three key factors: trust, privacy concerns, and perceived usefulness. This used to extract 18 observe variables in factor analysis. Following this
perspective aligns with previous studies (Adamson & Shine, 2003; Lucas analysis, we assessed the convergent validity using the AVE and CR
& Spider, 1999; Vijayasarathy, 2004) that have emphasized the essential metrics, which are shown in Table 1. Factor loadings ranged from 0.64
role of consumer perceptions as mediating variables in technology to 0.88, and all latent variables showed an AVE value greater than 0.5,
adoption. Expanding on this concept, Ahn et al. (2003) highlighted the indicating satisfactory convergence (Hair et al., 2017).
positive effects of online attributes, such as informational and system We used Cronbach’s alpha coefficient test (Cronbach, 1951) to
quality, on perceived usefulness and ease of use. Similarly, Sung et al. evaluate the survey questionnaire’s internal consistency reliability. Ac­
(2021) conducted research demonstrating that incorporating cording to Nunnally (1978), a reliability score of at least 0.70 is
AI-powered advanced technological capabilities significantly enhances considered acceptable. Table 1 shows that the latent variables have
the consumer experience, leading to positive attitudes and engagement. alpha values that fall between 0.826 and 0.914. The variables are being
Our current research, we draw upon these established connections and reliably measured by the questionnaire, as these values surpass the 0.7
introduce trust as a mediating factor. We believe that trust plays a threshold.
crucial role in connecting AI-driven personalization and user opinions, Table 2 presents the mean, standard deviation (SD), and a discrim­
serving as a driving force in the complex network of perceived user inant validity matrix. We applied the Fornell-Larcker criterion to assess
friendliness, perceived utility, and attitudes towards customer interac­ latent variables and mitigate potential multicollinearity concerns. Ac­
tion. Our research goal is to comprehend how AI-driven personalization cording to this criterion, the values on the upper right diagonal of the
influences customer engagement on social media platforms, with trust, matrix should surpass correlations with other variables, ideally equaling
privacy concern, and perceived usefulness as mediators. Therefore, hy­ or greater than the AVE’s square root. This process confirms the
potheses are proposed. discriminant validity of our constructed model (Fornell & Larcker,
1981).
H8. AI-enabled personalization and customer engagement are medi­
ated by customer trust perceptions. 4. Results
H9. AIenabled personalization and customer engagement are medi­
ated by customer privacy concern perceptions. 4.1. Sampling

H10. AI-enabled personalization and customer engagement are Table 3 presents the participant information, derived from survey
mediated by customer perceived usefulness. data that that were processed using statistical tools.
The sample for this study consisted of 245 respondents, comprising
3. Methods 148 males (60.41%) and 96 females (39.59%). A significant proportion
of the respondents were aged between 21 and 30 years olds (42%), had
3.1. Data collection procedure education levels below a bachelor’s degree or were college students
(77.87%), and earned a monthly salary of less than 10,000 baht. Most of
This research adopted a quantitative approach to gather primary the participants were Thai (78.27%) and reported spending over 5 h
data through survey questionnaires. Convenience sampling was used in daily on social media (40.57%). They used various social media plat­
the study in selecting participants. Google Forms facilitated the creation forms, including Facebook (80%), Instagram (60%), TikTok (56.55%),
of the survey questionnaire, with participants providing feedback using Line (45.49%), and WeChat (36.88%). Additionally, all participants had
a five-point rating scale. The majority of participants in this research experience with multiple online shopping platforms, such as Shopee
comprised lecturers and college students. A total of 245 questionnaires (74.59%) and Lazada (60.65%).
were returned via the online survey. There were 245 appropriate replies
considered for the final analysis. A sample size that is appropriate for the 4.2. Empirical testing of the hypothesized model
number of variables being examined is crucial for multivariate analysis.
According to Hair et al. (2006) and Sekaran and Bougie (2016) suggest The evaluation of model fit involved several key metrics, each
that a common range for the sample size is between 100 and 250 re­ associated with recommended thresholds (Hair et al., 2017). The
spondents. In this study, there were 5 constructs and 16 measuring model’s fit was evaluated using the following metrics: chi-squared ratio
items, indicating the minimum sample size should be at least 160 re­ to degrees of freedom (χ2/df ≤ 3), Root Mean Square Residual (RMR
spondents. This range ensures sufficient statistical power to detect sig­ ≤0.08), approximation’s root mean squared error (RMSEA ≤0.08),
nificant relationships among the variables in the construct model. comparative fit (CFI ≥0.9), incremental fit (IFI ≥0.9), Normed Fit (NFI
≥0.9), and Tucker Lewis (TLI ≥0.9) (Hair et al., 2017). The results are
3.2. Measurement scales displayed in Table 4, demonstrating the measurement model was in
good alignment with the data.
In this study, we examined 16 constructs derived from previously The results of the fitting test indicated that both the measurement

4
T. Teepapal Computers in Human Behavior 165 (2025) 108549

Table 1 likewise within an acceptable range. These models were appropriate for
Exploratory factor analysis and reliability analysis. further path analysis and satisfied the suggested fitting criteria. Table 4
Factors/Items Cronbach’s CR AVE Factor shows how well the structural model’s results fit as well. The model and
α Loading path coefficients are shown in Fig. 1, and the results show that re­
AI enable Personalization in social 0.882 0.80 0.57 ​ lationships are statistically significant with p-values less than 0.001.
media
Artificial intelligence (AI) systems on 0.757 4.3. Hypothesis testing
popular social media platforms
such as Line, Facebook, Instagram,
Twitter (X), and TikTok often
This study used the maximum likelihood method with analysis of
present content that matches your moment structures (AMOS) to assess the credibility of the proposed
personal needs and preferences. structural model. Hypothesis acceptance was determined by a C.R. (t-
Artificial intelligence (AI) technology 0.812 value) greater than 1.96 with a significance level below 0.05. Two
on social media platforms, they
separate tests were conducted to examine the relationships and media­
often present content, discounts, or
promotions that match your tion of the hypotheses. The first test assessed hypotheses H1 to H7 for
interests and personal preferences. direct relationships. The second test aimed to test the mediation hy­
Artificial intelligence (AI) systems 0.705 pothesis on H8 to H10 using a bootstrapping procedure and a user-
built into social media platforms
defined route in the estimand command.
improve your user experience.
Perception: Privacy concern 0.826 0.85 0.65 ​
Significant impacts are shown by the path analysis results, which are
You are unsure about the safety of 0.725 shown in Table 5. AI-enabled personalization exhibited a positive in­
storing personal information on fluence on trust perception (β = 0.58, t = 2.07, p < 0.001), privacy
social media platforms that use concern perception (β = 0.63, t = 9.02, p < 0.001), and perceived
artificial intelligence (AI) systems.
usefulness perception (β = 0.77, t = 12.28, p < 0.001), offering sub­
You are worried that your privacy 0.844
information will not be safe stantial support for H1, H2, and H3. Furthermore, the findings indicate
because artificial intelligence (AI) that trust perception significantly influences consumer engagement (β =
systems may access your 0.392, t = 4.16, p < 0.001), providing support for H4. However, the
information and use it without your
perception of privacy concerns did not significant impact consumer
permission.
You are concerned about protecting 0.855
engagement (β = 0.001, t = 0.10, p = 0.992), thus not supporting hy­
your information. When you know pothesis H5. Perceived usefulness was found to significantly affect
that there is an AI system on the consumer engagement (β = 0.392, t = 3.55, p < 0.001), supporting H6.
social media platforms you However, hypothesis 7 was not supported, suggesting that AI-enabled
regularly use
personalization did was not statistically significant customer engage­
Perceptions: Trust 0.853 0.81 0.59 ​
The social media platforms you are 0.742 ment (β = 0.176, t = 1.27, p = 0.20).
using are trusted and reliable
source of online information. 4.3.1. Mediation analysis
You believe that the content on social 0.887 To examine the mediation effect of consumer perceptions of trust on
media platforms is accurate.
Social media content can meet your 0.674
the relationship between AI-enabled personalization and customer
needs. engagement, a mediation analysis was conducted using 5000 boot­
Perception: Perceived usefulness 0.914 0.85 0.59 ​ strapping samples and a 95% bias-corrected confidence interval (CI).
Social media platforms have easy-to- 0.647 The analysis, detailed in Table 6, reveals a notable indirect effect on
use instructions that are
trust as a mediator between AI-enabled personalization and customer
customized to your interests.
Using social media platforms is 0.781 engagement (H8: β = 0.26, p = 0.03). Table 5 also shows that trust is
convenient and can be used strongly influenced by AI-enabled personalization (AI_P) (β = 0.72, p <
anywhere, anytime. 0.001), and that trust has a significant influence on customer engage­
Using social media platforms allows 0.846 ment (CE) (β = 0.39, p < 0.001). These results imply that the relation­
you to find a wider range of
information.
ship between AI-enabled personalization and customer engagement is
Using these social media platforms 0.800 fully mediated by trust, thus confirming the support for H8.
allows you to do many things Table 6 presents the results, which show that the indirect influence of
faster. AI-enabled personalization on customer engagement, mediated by
User engagement 0.829 0.69 0.57 ​
consumer privacy concerns, was not statistically significant (H9: β =
Social media platform content 0.808
stimulates your interest, that lead 0.001, p = 0.977). As shown in Table 5, the direct effect of AI-enabled
you to press Like and share the personalization (AI_P) had a significant influence on privacy concern
content. (β = 0.63, p < 0.001). On the other hand, the effect of privacy concern
You press Like on the social media 0.688 on customer engagement (CE) (β = 0.001, p = 0.992) was not statisti­
platform that matches your
preferences.
cally significant. These findings suggest that consumer perceptions of
You will recommend or share content 0.777 privacy concern only partially mediate the correlation among AI-
that matches your preferences with enabled personalization and engagement of customer. Therefore, H9
friends and family. was not supported.
In the relationship between AI-enabled personalization (AI_P) and
customer engagement (CE), the results presented in Table 6 indicate a
model and the structural model fit well; the measurement model’s chi-
significant indirect effect on perceived usefulness (H10: β = 0.23, p =
square ratio to degrees of freedom (CMIN/DF) was 2.16. Furthermore,
0.01). In addition, the direct effect of AI_P on perceived usefulness (β =
the recommended threshold of 0.9 was exceeded by the Normed Fit
0.77, p < 0.001), and the subsequent impact of perceived usefulness on
Index (NFI), Incremental Fit Index (IFI), Tucker-Lewis Index (TLI), and
CE (β = 0.33, p < 0.001), were both found to be statistically significant.
Comparative Fit Index (CFI) values, which were 0.927, 0.96, 0.939, and
These results indicate that consumer perceptions of perceived usefulness
0.9598, respectively. For the measurement model, the Root Mean
fully mediate the link between AI-enabled personalization and customer
Square Error of Approximation (RMSEA) value was 0.069, which is
engagement. Therefore, H10 was supported.

5
T. Teepapal Computers in Human Behavior 165 (2025) 108549

Table 2
Discriminant Validity (Fornell- Larcker criterion).
Factors Mean (SD) AI-enable Privacy Concern Trust Perceived usefulness Engagement

AI-enable Personalization 3.68 (1.08) 0.755 ​ ​ ​ ​

Perception: Privacy concern 3.72 (1.76) 0.611 0.806 ​ ​ ​
Perceptions: Trust 3.44 (0.98) 0.693 0.440 0.768 ​ ​
Perception: Perceived usefulness 3.93 (0.94) 0.736 0.548 0.684 0.768 ​
User Engagement 3.66 (1.01) 0.698 0.462 0.737 0.731 0.75498

Note: Diagonal elements in bold (AVE square roots).

Table 3
Descriptive statistics.
Items Variables Percentage Items Variables Percentage

Gender Male 60.25 Occupation College student 74.18

​ Female 39.75 ​ Employee 6.97
Nationality Thai 78.27 ​ Unemployed 18.85
​ China 18.45 Social media Facebook 80.00
​ Indonesia 3.28 Account Instagram (IG) 60.00
Age 18–20 36.10 ​ Twitter (X) 8.60
​ 21–30 42.20 ​ Lazada 60.65
​ 31–40 8.20 ​ Shopee 74.59
​ 41–50 9.90 ​ TikTok 56.55
​ More than 50 3.69 ​ WeChat 36.88
Education > bachelor’s degree 77.87 ​ QQ 16.39
​ Bachelor’s degree 12.30 ​ Line 45.49
​ < bachelor’s degree 9.83 Time spends >1 21.72
Salary (THB) > 10,000 60.30 on social media 1–2 29.10
​ 10,001–20,000 13.60 hour per day 3–4 8.61
​ 20,001–40,000 16.00 ​ >5 40.57
​ > 40,000 10.10 ​ ​ ​

Note: n = 245.

Table 4
Model fit indices.
Model χ2/df RMR RMSEA NFI IFI TLI CFI

Measurement model 2.167 2.331 0.049 0.055 0.069 0.074 0.927 0.918 0.96 0.95 0.939 0.939 0.959 0.951
Structural model

Note: Recommend index χ2/df ≤ 3, RMR ≤0.08, RMSEA ≤0.08, NFI ≥0.90, IFI ≥0.90, TLI ≥ 0.90, CFI ≥0.90.

Fig. 1. Structure model.

5. Discussions concerns, perceived usefulness, and consumer engagement. Our findings

highlight the intricate relationship between AI-driven personalization
This empirical study seeks to explore the influence of AI driven and consumer behavior, adding valuable insights to the current
personalization on businesses operating on social media platforms, literature.
focusing on its influence on customer perceptions of trust, privacy Our findings support the hypothesis that AI-enabled personalization

6
T. Teepapal Computers in Human Behavior 165 (2025) 108549

Table 5
Summary of finding.
Hypotheses Relationship β SE CR P-value Results

H1 Trust ← AI_P 0.72 0.064 10.66 *** Supported

H2 Privacy ← AI_P 0.63 0.072 9.02 *** Supported
H3 UseF ← AI_P 0.77 0.061 12.28 *** Supported
H4 CE ← Trust 0.39 0.092 4.16 *** Supported
H5 CE ← Privacy 0.001 0.066 0.010 0.992 Not supported
H6 CE ← UseF 0.336 0.088 3.55 *** Supported
H7 CE ← AI_P 0.176 0.126 1.27 0.202 Not supported

Note: AI_P = AI enable Personalization, Trust = Perception on trust, Privacy = Perception on privacy concern, UseF = Perception on perceived usefulness, CE =
Consumer Engagement, SE = Standardized Error, C.R. = Critical Ratio. ***p < 0.001, **p < 0.01, *p < 0.05.

Table 6
Findings from mediation analysis.
Hypotheses Relationship Indirect Effect SE Percentile Bootstrap 95% P-value Results

Lower Bound Upper Bound

H8 AI_P → Trust → CE 0.260** 0.86 0.108 0.451 0.003 Supported (Full Mediation)
H9 AI_P → Privacy → CE 0.001 0.68 − 0.138 0.135 0.977 Not Supported (No Mediation)
H10 AI_P → UseF → CE 0.236** 0.86 0.073 0.393 0.010 Supported (Full Mediation)

Note: n = 5000 (bootstrap), AI_P = AI enable Personalization, Trust = Perception on trust, Privacy = Perception on privacy concern, UseF = Perception on perceived
usefulness, CE = Consumer Engagement, SE = Standardized Error, C.R. = Critical Ratio, ***p < 0.001, **p < 0.01, *p < 0.05.

significantly affects consumer trust, privacy concerns, and perceived 6. Implications

usefulness, which aligns with previous studies by Bag et al. (2021) and
Krishen et al. (2021). Trust and perceived usefulness emerge as critical This study enhances our understanding of AI’s influence on con­
factors in influencing consumer engagement, underscoring their essen­ sumer engagement by examining how consumer perceptions mediate
tial role in how users interact with personalized content. While Gao and the relationship between AI-enabled personalization and customer
Liu (2022) identified privacy concerns as a significant factor impacting engagement. Trust and perceived usefulness have emerged as significant
consumer engagement, our study suggests otherwise. This discrepancy mediators in this relationship, highlighting the complexities of con­
could be attributed to differences in research methodologies or evolving sumer interactions with AI technologies. These findings contribute to
consumer attitudes towards privacy. This unexpected finding indicates both academic knowledge and practical applications by offering insights
that while privacy is a concern, it may not be as influential in shaping into the mediating factors and providing guidance for organizations
consumer behavior within the context of AI-driven personalization as seeking to leverage AI-enabled personalization to enhance consumer
previously thought. engagement.
Remarkably, the study finds that AI-driven personalization does not For academic.
exert a significant direct impact on customer engagement. Previous
studies, such as those by Alzaidi and Agag (2022), have suggested that 1. This research extends the Stimulus-Organism-Response (SOR) model
consumers generally trust products they discover on social media. by illustrating how AI stimuli impact consumer engagement,
However, our study highlights the complex nature of this relationship, providing a robust framework for future studies in AI and consumer
indicating that trust alone may not be sufficient to drive engagement. behavior contexts. The integration of AI as a stimulus provides a
This divergence may result from the specific nature of AI applications novel perspective, encouraging further exploration of AI’s role in
examined or variations in user behavior across different social media shaping consumer responses.
platforms. Future research should explore the relationship between 2. Recognizing trust and perceived usefulness as key mediators pro­
AI-driven personalization, consumer trust, and engagement across a vides valuable insights into the fundamental processes that influence
broader range of digital contexts. the relationship between AI-enabled personalization and customer
Privacy concerns pose a potential threat to the effectiveness of AI- engagement. These findings can inform future theoretical de­
enabled personalization by potentially deterring consumers from velopments by encouraging the exploration of additional mediators
sharing data necessary for personalized experiences (Aguirre, Mahr, or moderators that may further clarify consumer behavior in AI-
Grewal, de Ruyter, & Wetzels, 2015). Despite the lack of a clear corre­ driven scenarios.
lation between privacy concerns and engagement, it raises critical 3. The finding that privacy concerns do not significantly mediate the
questions about the personalization-privacy paradox, where consumers relationship between AI personalization and engagement challenges
may feel dissatisfied despite the perceived benefits of personalization existing assumptions. This opens new research avenues, encouraging
(Canhoto et al., 2023). This paradox presents a challenge for marketers, scholars to delve deeper into understanding privacy perceptions in
who must carefully balance offering valuable personalization with AI-enabled personalization and their varying impacts across different
respecting consumer privacy. demographic groups and cultural contexts.
Moreover, our study emphasizes the importance of perceived use­
fulness in driving consumer engagement (Liu & Tao, 2022). Customers’ For practitioners.
perceptions of the usefulness of AI-driven personalization greatly in­
fluence how they value and engage with it. However, the lack of a direct 1. Companies should prioritize building consumer trust and empha­
impact of AI-enabled personalization on customer engagement suggests sizing the perceived usefulness of AI applications in their e-com­
that other factors, such as content quality, user experience, or brand merce strategies. By enhancing these aspects, businesses can
loyalty, may play a more substantial role in fostering engagement. positively influence user experiences and foster stronger consumer

7
T. Teepapal Computers in Human Behavior 165 (2025) 108549

engagement, leading to increased brand loyalty and customer Ahn, T., Ryu, S., & Han, I. (2003). The impact of the online and offline features on the
user acceptance of internet shopping malls. Electronic Commerce Research and
retention.
Applications, 3(4), 405–420.
2. The study highlights a potential AI-personalization-privacy paradox, Alzaidi, M. S., & Agag, G. (2022). The role of trust and privacy concerns in using social
where efforts to personalize, content may inadvertently lead to pri­ media for e-retail services: The moderating role of COVID-19. Journal of Retailing and
vacy concerns. Practitioners must carefully balance personalization Consumer Services, 68, Article 103042.
Ameen, N., Tarhini, A., Reppe, A., & Anand, A. (2021). Customer experiences in the age
with robust privacy protection measures to avoid negative user ex­ of artificial intelligence. Computers in Human Behavior, 114, Article 106548.
periences and dissatisfaction. Maintaining this balance is essential to Andrzejak, E. G. (2023). AI-Powered digital transformation: Tools, benefits and
preserving customer confidence and averting possible objections. challenges for marketers – case study of LPP. Procedia Computer Science, 219,
397–404.
3. Recognizing the mediating effects of trust and perceived usefulness, Ashfaq, M., Yun, J., Yu, S., & Loureiro, S. (2020). I, chatbot: Modeling the determinants
businesses should adopt a customer-centric approach to AI imple­ of users’ satisfaction and continuance intention of AI-powered service agents.
mentation. This involves ensuring transparency in AI processes, Telematics and Informatics, 54, Article 101473. [Link]
tele.2020.101473
reliability in AI-generated content, and clear communication of the Bag, S., Sivarajah, U. S., Kumari, A., Giannakis, M., & Chowdhury, A. H. (2022). Journey
practical benefits of AI applications to users. By aligning AI strategies of customers in this digital era: Understanding the role of artificial intelligence
with consumer needs and expectations, companies can enhance technologies in user engagement and conversion. Benchmarking: An International
Journal, 29(7), 2074–2098.
engagement and drive long-term success. Campbell, C., Sands, S., Ferraro, C., Tsao, H. Y. J., & Mavrommatis, A. (2020). From data
to action: How marketers can leverage AI. Business Horizons, 63(2), 227–243.
Canhoto, A. I., Keegan, B. J., & Ryzhikh, M. (2023). Snakes and ladders: Unpacking the
personalisation-privacy paradox in the context of AI-enabled personalisation in the
6.1. Limitation physical retail environment. Information Systems Frontiers. [Link]
s10796-023-10369-7
This study acknowledges a number of limitations that may impact Chen, Y., & Zhang, Y. (2019). Understanding customer engagement in mobile social
commerce: A social influence perspective. International Journal of Information
the general interpretation and application of the findings. Addressing Management, 49, 366–377.
these constraints in future research could lead to a more comprehensive Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests.
understanding of AI-enabled personalization and its effects on consumer Psychometrika, 16, 297–334. [Link]
Davis, F. (1989). Perceived usefulness, perceived ease of use, and user acceptance of
behavior. First, the study’s time constraints resulted in a relatively small information technology. MIS Quarterly, 13, 319–340.
sample size, which may limit the generalizability of the results (Chen & Erragcha, N., & Romdhane, R. (2014). New faces of marketing in the era of the web:
Zhang, 2019). Moreover, the sample predominantly comprised univer­ From marketing 1.0 to marketing 3.0. Journal of Research in Marketing, 2(2),
137–142.
sity students aged 18–22 years, which limited the results’ generaliz­ Fornell, C., & Larcker, D. F. (1981). Structural equation models with unobservable
ability. In order to enhance both the validity and significance of the variables and measurement error: Algebra and statistics. Journal of Marketing
results, future studies ought to aim to incorporate a more extensive and Research, 18(3), 382–388.
Gao, L., Li, G., Tsai, F., Gao, C., Zhu, M., & Qu, X. (2023). The impact of artificial
diverse group of participants. Second, this study exclusively examines
intelligence stimuli on customer engagement and value co-creation: The moderating
consumer perceptions related to trust, privacy concerns, and perceived role of customer ability readiness. The Journal of Research in Indian Medicine, 17(2),
usefulness. While these factors are critical, other psychological variables 317–333.
Gao, Y., & Liu, H. (2022). Artificial intelligence-enabled personalization in interactive
such as motivation, learning, and lifestyle (Li & Wang, 2018) also play
marketing: A customer journey perspective. The Journal of Research in Indian
significant roles in shaping consumer behavior. Future studies could Medicine, 2040–7122.
explore these additional factors to provide a more holistic understanding Gefen, D., & Straub, D. (2003). Managing user trust in B2C e-services. E-Service Journal, 2
of consumer engagement in AI-driven environments. Third, the (2), 7–24.
Gołąb-Andrzejak, E. (2023). AI-Powered digital transformation: Tools, benefits and
cross-sectional nature of this study limits its ability to capture the challenges for marketers–case study of LPP. Procedia Computer Science, (219),
long-term effects of AI-enabled personalization on consumer engage­ 397–404.
ment. To gain deeper insights, future research should consider longitu­ Gutierrez, N., Freundschuh, S., & Ahn, G. (2019). From privacy to control:
Reconceptualizing a complex construct in an age of information disclosure. Journal
dinal studies that observe user behavior over an extended period. This of the Association for Information Science and Technology, 70(10), 1090–1100.
approach could reveal causal relationships and offer a better under­ Ha, S., & Stoel, L. (2009). Consumer e-shopping acceptance: Antecedents in a technology
standing of how social media platforms and AI technologies evolve and acceptance model. Journal of Business Research, 62, 565–571.
Hair, J., Black, W., Babin, B., Anderson, R., & Tatham, R. (2006). Multivariate data
impact marketing strategies over time. Finally, while this study provides analysis (6th ed.). Upper Saddle River, NJ: Pearson Prentice Hall.
valuable insights into the general impact of AI-enabled personalization, Hair, J. F., Hult, G. T. M., Ringle, C. M., & Sarstedt, M. (2017). A primer on partial least
the effects may vary across different social media platforms. Future squares structural equation modeling (PLS-SEM) (2nd ed.). Thousand Oaks, CA: Sage
Publications Inc.
research could focus on platform-specific studies to uncover unique
Haleem, A. J., Mohd, M., Qadri, A., Singh, R. P., & Suman, R. (2022). Artificial
dynamics and tailor marketing strategies to the characteristics of each intelligence (AI) applications for marketing: A literature-based study. International
platform. Journal of Intelligent Networks, 3, 119–132.
Hollebeek, L., Glynn, M., & Brodie, R. (2014). Consumer brand engagement in social
media: Conceptualization, scale development and validation. Journal of Interactive
Declaration of competing interest Marketing, 28. [Link]
Huang, R., Kim, M., & Lennon, S. (2022). Trust as a second-order construct: Investigating
the relationship between consumers and virtual agents. Telematics and Informatics, 70
The authors declare that they have no known competing financial (101811), 1–12.
interests or personal relationships that could have appeared to influence Huang, M. H., & Rust, R. T. (2020). A strategic framework for artificial intelligence in
the work reported in this paper. marketing. Journal of the Academy of Marketing Science, 49(1), 30–50.
Jin, B. E., & Shin, D. C. (2020). Changing the game to compete: Innovations in the
fashion retail industry from the disruptive business model. Business Horizons, 63(3),
Data availability 301–311.
Kannan, P. K. (2017). Digital marketing: A framework, review and research agenda.
International Journal of Research in Marketing, 34(1), 22–45.
The data that has been used is confidential. Kaplan, A., & Haenlein, M. (2019). Siri, siri, in my hand: Who’s the fairest in the land?
On the interpretations, illustrations, and implications of artificial intelligence.
References Business Horizons, 62(1), 15–25.
Krishen, A. S., Dwivedi, Y. K., Bindu, N., & Kumar, K. S. (2021). A broad overview of
interactive digital marketing: A bibliometric network analysis. Journal of Business
Adamson, I., & Shine, J. (2003). Extending the new technology acceptance model to
Research, 131, 183–195.
measure end user information systems satisfaction in a mandatory environment: A
Kumar, A., & Kalse, A. (2021). Usage and adoption of artificial intelligence in SMEs.
bank’s treasury. Technology Analysis & Strategic Management, 15(4), 441–455.
Materials Today: Proceedings.
Aguirre, E., Mahr, D., Grewal, D., de Ruyter, K., & Wetzels, M. (2015). Unraveling the
personalization paradox: The effect of information collection and trust-building
strategies on online advertisement effectiveness. Journal of Retailing, 91(1), 34–49.

8
T. Teepapal Computers in Human Behavior 165 (2025) 108549

Lalicic, L., & Weismayer, C. (2021). Consumers’ reasons and perceived value co-creation Roggeveen, A. L., Grewal, D., Karsberg, J., Noble, S. M., Nordfält, J., Patrick, V. M., &
of using artificial intelligence-enabled travel service agents. Journal of Business Olson, R. (2021). Forging meaningful consumer-brand relationships through creative
Research, 129, 891–901. merchandise offerings and innovative merchandising strategies. Journal of Retailing,
Li, X. (2019). Understanding Chinese consumers’ trust in social commerce: An 97(1), 81–98.
exploratory study. Information Systems Frontiers, 21(1), 229–243. Sadriwala, M. F., & Sadriwala, K. F. (2022). Perceived usefulness and ease of use of
Li, H., & Wang, Y. (2018). Understanding the influence of hedonic motivation on artificial intelligence on marketing innovation. International Journal of Innovation in
consumer engagement in social commerce: An empirical study. Information & the Digital Economy, 13(3), 10.
Management, 55(8), 980–991. Sakas, D. P., Reklitisa, D. P., Giannakopoulosa, N. T., & Trivellas, P. (2023). The
Liu, K., & Tao, D. (2022). The roles of trust, personalization, loss of privacy, and influence of websites user engagement on the development of digital competitive
anthropomorphism in public acceptance of smart healthcare services. Computers in advantage and digital brand name in logistics startups. European research on
Human Behavior, 127(8), Article 107026. management and business economics, 29(100221), 1–18.
Lucas, J. H. C., & Spider, V. K. (1999). Technology use and performance: A field study of Sekaran, U., & Bougie, R. (2016). Research methods for business: A skill-building approach
broker workstations. Decision Sciences, 30(2), 291–310. (7th ed.). West Sussex: Wiley & Sons.
Ma, L., & Sun, B. (2020). Machine learning and AI in marketing – connecting computing Sharma, S., Singh, S., Kujur, F., & Das, G. (2021). Social media activities and its influence
power to human insight. International Journal of Research in Marketing, 37(3), on customer-brand relationship: An empirical study of apparel retailers’ activity in
481–504. India. Journal of Theoretical and Applied Electronic Commerce Research, 16(4),
Maduku, D. K., Mpinganjira, M., Rana, N. P., Thusi, P., Ledikwe, A., & Mkhize, N. H. 602–617.
(2023). Assessing customer passion, commitment, and word-of-mouth intentions in So, K. K. F., Wei, W., & Martin, D. (2021). Understanding customer engagement and
digital assistant usage: The moderating role of technology anxiety. Journal of social media activities in tourism: A latent profile analysis and cross-validatio.
Retailing and Consumer Services, 71(103208), 1–13. Journal of Business Research, 129, 474–483.
Mehrabian, A., & Russell, J. A. (1974). An approach to environmental psychology. MIT Sung, C., Bae, S., Han, D., & Kwon, O. (2021). Consumer engagement via interactive
Press. artificial intelligence and mixed reality. International Journal of Information
Na, S., Heo, S., Han, S., Shin, Y., & Roh, Y. (2022). Acceptance model of artificial Management, 60, Article 102382.
intelligence (AI)-Based technologies in construction firms: Applying the technology Swani, K., Milne, G. R., & Slepchuk, A. N. (2021). Revisiting trust and privacy concern in
acceptance model (TAM) in combination with the technology–organisation– consumers’ perceptions of marketing information management practices:
environment (TOE) framework. Buildings, 12(90), 1–17. Replication and extension. Journal of Interactive Marketing, 56, 137–158.
Pantano, E., & Vannucci, V. (2019). Who is innovating? An exploratory research of Van Der Heijden, H., Verhagen, T., & Creemers, M. (2003). Understanding online.
digital technologies diffusion in retail industry. Journal of Retailing and Consumer Purchase intentions: Contributions from technology and trust perspectives. European
Services, 49, 297–304. Journal of Information Systems, 12(1), 41–48.
Payne, E. M., Dahl, A. J., & Peltier, J. (2021a). Digital servitization value co-creation Vijayasarathy, L. R. (2004). Predicting consumer intentions to use on-line shopping: The
framework for AI services: A research agenda for digital transformation in financial case for an augmented technology acceptance model. Information & Management, 41
service ecosystems. The Journal of Research in Indian Medicine, 15(2), 200–222. (6), 747–776.
Perez-Vega, R., Kaartemo, V., Lages, C. R., Razavi, N. B., & M€annist€o, J. (2021).
Reshaping the contexts of online customer engagement behavior via artificial
intelligence: A conceptual framework. Journal of Business Research, 129, 902–910.

9
 Update
Computers in Human Behavior
Volume 174, Issue , January 2026, Page

DOI: [Link]
 Computers in Human Behavior 174 (2026) 108797

Contents lists available at ScienceDirect

Computers in Human Behavior

journal homepage: [Link]/locate/comphumbeh

Corrigendum to “AI-driven personalization: Unraveling consumer

perceptions in social media engagement” [Computers in Human Behavior
165 (2025) 108549]
Tanawat Teepapal
International College, Rajamangala University of Technology Krungthep, Bangkok, Thailand

The author would like to inform readers of the following change in [Link]/10.1080/10864415.2019.1619907
the reference list of this manuscript: The author would also like to add the following declaration:
Li, X. (2019). Understanding Chinese consumers’ trust in social com­ Declaration of Generative AI in Scientific Writing
merce: An exploratory study. Information Systems Frontiers, 21(1), During the preparation of this work, the author used ChatGPT
229–243 (OpenAI) for language editing and improving readability. After using
Should be replaced by this tool, the author reviewed and edited the content as needed and take
Lin, X., Wang, X., & Hajli, N. (2019). Building E-commerce satisfaction full responsibility for the final version of the manuscript.
and boosting sales: The role of social commerce trust and its antecedents. The author would like to apologize for any inconvenience caused.
International Journal of Electronic Commerce, 23(3), 328–363. https://

DOI of original article: [Link]

E-mail address: tanawat.t@[Link].

[Link]

Available online 13 September 2025

0747-5632/© 2025 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.