Negative Effectsof Revealing AIInvolvementin Products Mediationby Authenticityand Risk Moderationby Trustin AIand Familiaritywith AI
Negative Effectsof Revealing AIInvolvementin Products Mediationby Authenticityand Risk Moderationby Trustin AIand Familiaritywith AI
net/publication/383206824
Article in International Journal of Intelligent Systems and Applications in Engineering · July 2024
DOI: 10.5281/zenodo.13337499
CITATIONS READS
0 198
1 author:
Gedas Kučinskas
ISM University of Management and Economics
11 PUBLICATIONS 20 CITATIONS
SEE PROFILE
All content following this page was uploaded by Gedas Kučinskas on 18 August 2024.
Keywords: Artificial intelligence, consumer perceptions, product authorship, authenticity, purchase decisions.
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1530
where originality is highly valued. Perceived risk involves These AI-driven tools streamline demand forecasting,
uncertainties consumers may feel about the quality and inventory control, and logistics, ensuring more efficient
personal relevance of products designed by algorithms. operations and reduced operational costs (Richey et al.,
This study also investigates how trust in AI and familiarity 2023). Additionally, AI facilitates more personalized
with AI-generated content can moderate these effects, customer interactions through advanced chatbots and
potentially enhancing consumer acceptance and purchase virtual assistants, significantly improving customer
intentions. engagement (Rane, 2023). However, the adoption of AI is
fraught with challenges such as data privacy, ethical
The hypotheses guiding this research propose that while AI
concerns, and the complexities of system integration,
involvement might initially increase perceived risks and
necessitating meticulous management strategies to
diminish perceptions of authenticity, thus potentially
overcome these barriers (Brendel et al., 2021). Despite
dampening consumer enthusiasm, higher levels of trust in
these challenges, AI's potential to enhance growth,
AI and increased familiarity with AI-generated content
competitiveness, and sustainability in the consumer
could mitigate these effects. These factors may help foster
products sector is substantial, promising major advantages
more favorable attitudes towards AI-generated products,
for companies that can effectively harness its capabilities
leading to greater consumer willingness to purchase and
(Wamba-Taguimdje et al., 2020).
engage with these products.
Comprehending consumer perceptions of AI-generated
By addressing this research gap, the study aims to provide
products is pivotal for fostering perceived value, brand
deeper insights into the dynamics of consumer acceptance
loyalty, guiding product development, and securing market
of AI-generated products. These insights will be invaluable
acceptance. The way consumers perceive AI-integrated
for marketers and product developers seeking to effectively
products influences not only individual purchasing
integrate AI into product design and marketing strategies,
behaviors but also shapes broader market trends (Marti et
ensuring that AI-driven innovations resonate positively
al., 2024). Overcoming the barriers to AI related products
with consumers and lead to successful market outcomes.
adoption and enhancing consumer trust can significantly
This research is particularly relevant in the context of elevate businesses innovation and competitiveness.
today's rapidly digitizing economy where AI's capabilities
Consumer acceptance and resistance to new technological
are continuously expanding. By focusing on AI-generated
products are shaped by a variety of factors that influence
products, this study addresses a critical aspect of AI's role
their perceptions and adoption behaviors. Technological
in modern consumer markets—its impact on consumer
readiness does not directly impact consumer attitudes but
perception and decision-making. The scientific
instead influences them indirectly through characteristics
contribution of this study lies in its exploration of how AI
associated with perceived innovation. Factors such as ease
involvement influences consumer perceptions of
of use, efficiency (Verhoef et al., 2017), and the latest
authenticity and risk, areas that have not been extensively
technological features like AI and IoT connectivity attract
studied in the context of AI-generated consumer products.
many consumers, while concerns about privacy, usability,
Furthermore, by examining the moderating roles of trust in
cost, and a lack of trust due to past negative experiences or
AI and familiarity with AI-generated content, this study
deceptive marketing deter others (De Bruyn et al., 2020).
contributes to the broader discourse on technology
Consumer opinions are also shaped by endorsements from
acceptance models, offering insights that could help
influencers, peer recommendations, and positive reviews,
businesses better integrate AI in ways that align with
which can significantly influence the wider acceptance of
consumer expectations and enhance market acceptance.
new technologies. Moreover, AI impacted, clear value
2. Literature Review & Hypothesis Buildup propositions and intuitive user interfaces are crucial in
The rapid evolution of e-commerce has led to an enhancing product acceptance, making it essential for
exponential increase in data volume, overtaking traditional companies to address issues related to privacy,
statistical methods and fundamentally transforming affordability, and usability to widen technology adoption
marketing strategies with the advent of AI-driven (Ismatullaev & Kim, 2024; Sohn & Kwon, 2020).
predictive analytics like data mining and machine learning The psychological impact of technology on consumer
(Gupta et al., 2023). Artificial intelligence (AI) has perceptions significantly influences how they value and
revolutionized the consumer products industry by assess product quality. Innovations are perceived as more
integrating sophisticated data analysis capabilities, valuable when they are seen as cutting-edge and superior.
including computer vision, natural language processing, Cognitive biases such as the "halo effect" can enhance
and machine learning, reshaping how companies approach these perceptions, emphasizing even minor technological
product development, supply chain optimization, and advancements. The added functionality and potential for
customer experience enhancement (Haleem et al., 2022). personalization enhance the appeal of these products,
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1531
potentially increasing their perceived status and implying a (Kang & Kim, 2013).
higher social prestige. However, the complexity inherent in
Perceived Performance Risk involves the uncertainty about
new technological products can also invoke concerns about
a product's functionality and its ability to perform as
their reliability and compatibility, possibly detracting from
expected. In the realm of AI-driven products, this risk is
their perceived value and quality (Lim & Zhang, 2022;
exacerbated by the complex nature of the technologies
Straub, 2009; Yang et al., 2016).
involved and the consumer's inability to fully test or
The integration of AI in consumer products is not only a understand the product prior to purchase. Performance risk
question of technological advancement but also deeply tied is a crucial consideration in industries like automotive or
to how authenticity and perceived risks—financial and electronics, where a failure to meet performance
performance—are managed. Authenticity (Newman & expectations can lead to dissatisfaction and significant
Bloom, 2012) in AI products extends beyond mere consumer regret (Sääksjärvi & Lampinen, 2005; Stampfl,
physical attributes; it encompasses the reliability, ethical 1978). Empirical findings suggest that performance risk
standards, and transparency that come with AI integrations. and financial risk can be mitigated by enhancing perceived
The acceptance and adoption of AI technologies can be quality and reducing perceived sacrifices. Strategies that
significantly hindered by perceived risks, whether elevate perceived quality, such as leveraging strong brand
financial, such as the potential for financial loss or names or trustworthy store environments, can significantly
unforeseen expenses, or performance-related, such as lower both types of perceived risks, thereby increasing the
doubts about product reliability or compatibility (Agarwal perceived value of the products (Agarwal & Teas, 2001).
& Teas, 2001; Bauer, 1960; Ha, 2006).
Furthermore, the role of consumer innovativeness and its
Perceived authenticity is the degree to which consumers impact on the perception of these risks illustrates that
regard a product as genuine, accurate, and consistent with familiarity with the technology can reduce the perceived
its intended purpose and design. This notion is closely risk, encouraging adoption. Consumers familiar with a
linked to brand authenticity, which assesses whether product category typically show lower perceived financial
consumers view a brand as authentic, reliable, and and performance risks, which boosts their willingness to
supportive. The importance of perceived product purchase new and innovative products (Hirunyawipada &
authenticity lies in its ability to impact consumer attitudes Paswan, 2006).
and behaviors towards a product. Products deemed
Familiarity with AI technologies significantly influences
authentic are often linked to greater perceived value and
consumer acceptance. Regular exposure to and interaction
trust in the brand, which can foster positive consumer
with AI-enhanced products reduce perceived risks and
actions (Nunes et al., 2021). This includes heightened
build trust, making these technologies less intimidating and
loyalty and a greater willingness to pay a premium for the
more accessible to consumers. Empirical studies support
product (de Kerviler et al., 2021; Newman & Bloom,
the idea that familiarity with AI not only enhances its
2012).
perceived utility and ease of use but also fosters a more
Perceived Financial Risk is particularly poignant in positive attitude towards its adoption. Tailored educational
consumer decision-making, encapsulating the potential initiatives and hands-on experiences with AI can further
negative financial outcomes from a purchase. This type of reduce consumer apprehensions and enhance their comfort
risk becomes more pronounced in scenarios involving and trust in these technologies (Horowitz et al., 2023;
high-cost items or new product categories where the Kelly et al., 2023; Wanner et al., 2022).
financial stakes and the potential for regret are substantial.
Providing detailed product information plays a crucial role
For instance, consumers might worry about the
in empowering consumers, enabling them to make
authenticity of a luxury item or the reliability of a new tech
informed decisions and enhancing their overall satisfaction
gadget, which could lead to significant financial loss if the
with AI-enhanced products. Transparency about product
product fails to meet expectations (Aldousari et al., 2017;
features, functionality, and the underlying AI mechanisms
Ha, 2006) (Aldousari et al., 2017; Vlaev et al., 2009).
reassures consumers, mitigating fears related to AI
Research has highlighted various dimensions of perceived
involvement and fostering a sense of control over their
risks that affect consumer attitudes towards products,
purchasing decisions. Clear, comprehensive information
including store brands or environmentally sustainable
about how AI technology’s function and their benefits can
apparel. Studies have shown that financial risk can deter
significantly alleviate consumer concerns, boosting trust
consumers from purchasing due to fears of poor quality or
and loyalty (Dhagarra et al., 2020; Sangtani & Murshed,
additional costs that may arise after the purchase.
2017; Sun et al., 2023).
Addressing these financial risks through clear
communication and assurances can enhance consumer trust The visibility of product outcomes significantly affects
and willingness to engage with new product categories consumer perceptions of risk and confidence.
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1532
Transparency in demonstrating how products operate, 3. Hypotheses Testing
perform, and deliver value reduces perceived risks
This study comprised three experiments aimed at
associated with ambiguity and uncertainty, enhancing
understanding the influence of AI involvement and product
consumer trust and satisfaction. The ability for consumers
type on consumer perceptions and behaviors: Study 1
to directly observe and evaluate the outcomes of using a
aimed to explore how different products (book vs. canvas
product reassures them of its reliability, efficacy, and
print) and the presence of AI involvement (hinted or not)
quality, fostering greater confidence and decision-making
influence consumer perceptions and behaviors.
security (Betzing et al., 2020; Kim & Jin, 2018; Reck et
Specifically, the research sought to understand the impact
al., 2022).
of these variables on a range of consumer responses,
Building on the previous research, the introduction of including valuation, purchasing intent, and subjective
Artificial Intelligence (AI) in product creation poses assessments of product features and potential risks.
significant implications for consumer perceptions, Building upon findings of study 1, Study 2 explored
particularly concerning authenticity and perceived risks. whether including additional information—such as
Consumers might view AI-generated products as less potential access to a detailed summary, table of contents,
authentic, potentially due to the perceived lack of human sample chapters from various parts of the book, and
involvement, which is traditionally associated with several positive reviews—affects the perceived risks
creativity and quality. Moreover, AI involvement might associated with AI-created products, in the e-book context.
elevate perceived risks—both performance and financial— Given the limited success of information provision, study 3
due to uncertainties about the product's reliability and aimed to investigate how limiting the visibility of the final
value. However, these negative perceptions might be product outcome, in this case canvas print, may affect
moderated by factors such as trust in AI and familiarity consumer perceptions. As all studies had factorial design
with AI, which can alleviate concerns and promote Analysis of Variance (ANOVA), were employed to assess
acceptance. Detailed product information and transparency the main effects and interactions of the variables on
regarding AI's role in product outcomes might mitigate consumer perceptions.
perceived risks and enhance consumer willingness to
Participants and Design. For all studies participants were
engage with AI-generated products. The set of hypotheses
recruited using MTurk, with Cloudresearch controls. Study
are proposed:
1 had 298 (Mage = 39.99, SDage = 11.89, women = 48.32%)
H1: AI involvement cues increase perceived risks of participants and employed a 2x2 factorial design to
products. investigate how customer perceptions were impacted by
product (book vs. canvas print) and AI involvement (no AI
H2: AI involvement cues negatively influence consumer
hint vs. AI hint). The study 2 involved 126 (M age = 41.02,
perceptions of product authenticity.
SDage = 11.79, women=53.17%) and structured as a 2x2
H3: AI involvement cues negatively influence consumer factorial arrangement examining the influence of
willingness to buy products. authorship (human vs. AI) and varying levels of
H4: Higher levels of trust in AI positively moderate the information on outcome (low vs. high) on consumer
relationship between AI involvement and consumer perceptions. Study 3, a total of 202 (M age = 42.1 SDage =
attitudes towards AI-generated products, as well as 12.1, women = 53.5%) participants and 2x2 factorial
purchase intentions. design to investigate how customer perceptions were
impacted by AI engagement (artist vs. AI) and different
H5: Increased familiarity with AI-generated content canvas print output visibility levels (visible vs. blurry).
mitigates the negative effects of AI involvement cues on
consumer perceptions and purchase intentions. Measures. In study 1, a variety of consumer behavior
measures were utilized in this study, all employing a 7-
H6: Providing detailed product information alters point Likert scale (Sullivan & Artino, 2013). The measures
perceived risks associated with these products. assessed a broad spectrum of consumer attitudes and
H7: Perceived risks negatively mediate the relationship perceptions regarding products, including willingness to
between AI involvement and purchase intentions. pay, likelihood to buy, perceived financial and
performance risks (Biswas & Biswas, 2004) (the higher the
H8: Perceived authenticity negatively mediates the rating – the higher the risk), perceived cost of production,
relationship between AI involvement and purchase and perceptions of uniqueness, authenticity, originality,
intentions. and psychological ownership (Atasoy & Morewedge,
H9: The visibility of the final product outcome significantly 2018). Measures of perceived effort in creation
reduces perceived risks of AI-generated products and permanence, credibility of the author, overall quality, and
increases purchase intentions. general product attitudes (good, like, interesting, α=0.92)
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1533
were also included. Participants were also asked about effects of perceived artificial intelligence (AI) involvement
their familiarity with AI-generated content and their trust on consumer behaviour, a significant difference across
in artificial intelligence. AI involvement impact assessed various measures was noted through an analysis of
through manipulation checks comparing conditions with variance (ANOVA). Notably, the willingness to pay was
and without AI hints (no AI hint vs. AI hint). Study 2 and higher in conditions without an AI hint (M _noAI=21.65,
Study 3 included less but key concepts - willingness to SD=16.09 vs. M_AI=16.21, SD=18.67, F(1,296)=7.22,
pay, likelihood to buy, perceived financial and p<0.01, η²=0.02). This trend was mirrored in the
performance risks, perceptions of authenticity, quality, willingness to buy (M_noAI=3.45, SD=1.92 vs. M_AI=2.89,
enjoyment, effort. SD=1.97, F(1,296)=6.24, p<0.01, η²=0.02).
Manipulation checks. In Study 1, manipulation checks Similarly, perceptions of financial and performance risk
assessed the impact of AI involvement with and without AI were significantly lower in the non-AI hinted conditions,
hints. The results indicated successful manipulation, with indicating a lower risk perception associated with non-AI
significantly higher perceptions of human involvement in products (financial risk: M_noAI=2.15, SD=1.59 vs.
the no AI hint condition compared to the AI hint condition M_AI=2.58, SD=1.83, F(1,296)=4.69, p<0.05, η²=0.02;
(M_noAi=5.059, SD=1.497 vs M_AI=2.116, SD=1.791, performance risk: M_noAI=3.93, SD=1.73 vs. M_AI=4.49,
F(1,296)=235.88, p<0.001, η²=.45). In Study 2, two SD=2.01, F(1,296)=6.58, p<0.05, η²=0.02). Authenticity was
manipulation checks were conducted. Firstly, perception of significantly higher without the AI hint (M_noAI=4.34,
human involvement was higher without AI author hints SD=1.61 vs. M_AI=3.56, SD=1.77, F(1,296)=15.55, p<0.001,
(M_human=5.734, SD=1.13) than with AI hints η²=0.05). This observation extended to originality
(M_human=5.734, SD=1.13 vs M_AI=2.419, SD=1.912, (M_noAI=4.12, SD=1.73 vs. M_AI=3.51, SD=1.80,
F(1,124)=139.19, p<.00001, η²=.59. Secondly, perceptions of F(1,294)=8.74, p<0.01, η²=0.03) and the feeling of
AI involvement were higher without human authorship distinctiveness (M_noAI=3.51, SD=1.74 vs. M_AI=2.90,
hints (M_human=2.656, SD=1.535 vs M_AI=6.323, SD=1.81, F(1,296)=8.77, p<0.01, η²=0.03). A significantly
SD=0.971, F(1,124)=258.34, p<.00001, η²=.71. Both results greater perceived effort was associated with products not
indicated successful manipulation of authorship hints. In hinted as AI-created (M_noAI=4.40, SD=1.55 vs.
Study 3, manipulation checks confirmed the perception of M_AI=3.11, SD=1.79, F(1,296)=44.34, p<0.001, η²=0.13).
AI involvement being higher with AI hints (M _artist=4.776, Similarly, perceptions of permanence (M_noAI=4.59,
SD=1.643 vs M_AI=4.776, SD=2.003), F(1,200)=54.55, SD=1.60 vs. M_AI=3.76, SD=1.80, F(1,296)=17.51, p<0.001,
p<.00001, η²=.22. Additionally, outcome visibility η²=0.06) and credibility (M_noAI=4.40, SD=1.37 vs.
perception was clearer when considering canvas prints M_AI=3.63, SD=1.74, F(1,296)=17.72, p<0.001, η²=0.06)
with visible outcomes (M_visible=5.919, SD=0.829 vs were significantly higher without AI involvement. Overall
M_blurred=2.864, SD=1.831, F(1,200)= 236.27, p<.00001, product quality perceptions were also notably higher
η²=.62) indicating successful manipulation of both AI (M_noAI=3.80, SD=1.26 vs. M_AI=3.25, SD=1.28,
author hints and outcome visibility perceptions in Study3. F(1,296)=13.92, p<0.001, η²=0.05), as were general product
attitudes (M_noAI=4.56, SD=1.51 vs. M_AI=3.86, SD=1.65,
3.1. Study 1: Impact of AI Involvement and Product
F(1,296)=14.72, p<0.001, η²=0.05). These findings
Type on Consumer Perceptions and Purchase
underscore a generally more positive consumer response to
Intentions
products when there is no suggestion of AI involvement
Main effects of AI involvement hint. In examining the (Table 1).
Table 1. No AI involvement hint vs disclosed AI involvement main effects
Variable Mean_noAI SD_noAI Mean_AI SD_AI F_statistic p_value η2
Willingness to pay 21.65 16.09 16.21 18.67 7.22 0.00762 0.02
Willingness to buy 3.45 1.92 2.89 1.97 6.24 0.01301 0.02
Financial risk 2.15 1.59 2.58 1.83 4.69 0.03114 0.02
Performance risk 3.93 1.73 4.49 2.01 6.58 0.01079 0.02
Perceived costs 13.74 14.68 12.10 17.37 0.77 0.38000 0
Unique 3.68 1.70 3.32 1.78 3.12 0.07850 0.01
Authentic 4.34 1.61 3.56 1.77 15.55 <0.001 0.05
Original 4.12 1.73 3.51 1.80 8.74 0.00336 0.03
Distinctiveness 3.51 1.74 2.90 1.81 8.77 0.00331 0.03
Perceived effort 4.40 1.55 3.11 1.79 44.34 <0.001 0.13
Ownership 3.72 1.97 3.31 1.99 3.25 0.07262 0.01
Enjoyment 4.34 1.64 4.05 1.82 2.05 0.15356 0.01
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1534
Variable Mean_noAI SD_noAI Mean_AI SD_AI F_statistic p_value η2
Permanence 4.59 1.60 3.76 1.80 17.51 <0.001 0.06
Credibility 4.40 1.37 3.63 1.74 17.72 <0.001 0.06
Quality 3.80 1.26 3.25 1.28 13.92 <0.001 0.05
Product attitudes 4.56 1.51 3.86 1.65 14.72 <0.001 0.05
Easiness to copy 4.50 1.51 4.77 1.50 2.46 0.11749 0.01
Main effects of product form (book vs print canvas). In a Enjoyment levels (F(1,294) = 4.05, p = .0345) were notably
comparative analysis of variance (ANOVA) involving different for canvas prints with an AI hint compared to
participants exposed to books versus print canvas, those without, suggesting that AI hints may adversely
significant disparities were observed across multiple affect the emotional response to art-based products (diff = -
measures. Participants exhibited a higher willingness to 0.739, p = .036, M_noAI_canvas=4.36 vs M_AI_canvas=4.52).
pay for the print canvas (M_book=11.86, SD=9.99 vs. Permanence perception (F(1,294) = 4.058, p = .0449) was
M_canvas=26.11, SD=20.45, F(1,296)=58.38, p<0.001, significantly lower for canvas prints with an AI hint
η²=0.21) and a lower perception of performance risk compared to those without, suggesting that AI involvement
(M_book=4.85, SD=1.75 vs. M_canvas=3.55, SD=1.81, might influence perceived durability or longevity of art-
F(1,296)=39.96, p<0.001, η²=0.12). based products (diff = -0.782, p = .023, M_noAI_canvas=4.59
vs M_AI_canvas=3.35). For books, although there was a
Additionally, the perceived cost was higher for the print
noticeable difference in perceived permanence between
canvas (M_book=8.74, SD=10.86 vs. M_canvas=17.12,
conditions with and without AI hints, it was not as
SD=19.07, F(1,296)=21.76, p<0.001, η²=0.08). Quality
pronounced as in the case of canvas prints (diff = -0.453, p
perceptions were also more favorable for the print canvas
= .3550, M_noAI_Book=4.58 vs M_AI_Book=4.13).
(M_book=3.97, SD=1.59 vs. M_canvas=3.09, SD=0.68,
F(1,296)=38.50, p<0.001, η²=0.16), as was enjoyment Credibility was significantly affected (F(1,294) = 10.508, p =
(M_book=3.95, SD=1.77 vs. M_canvas=4.44, SD=1.67, .0013) by AI hints in both product forms. For books,
F(1,296)=5.88, p=0.016, η²=0.02). Participants perceived credibility was notably lower when there was an AI hint
more effort in the creation of the print canvas product (diff = -0.672, p = .0328, M_noAI_Book=4.75 vs
(M_book=3.99, SD=1.70 vs. M_canvas=3.54, SD=1.85, M_AI_Book=3.40). Similarly, for canvas prints, credibility
F(1,296)=4.77, p=0.03, η²=0.02), and found it more was reduced with AI involvement, although the effect was
challenging to copy (M_book=4.45, SD=1.45 vs. not significant (diff = -0.431, p = .4232, M_noAI_canvas=4.08
M_canvas=4.82, SD=1.56, F(1,296)=4.50, p=0.035, η²=0.02). vs M_AI_canvas=3.88).
However, variables like willingness to buy, financial risk, Quality perceptions differed significantly (F(1,294) = 18.071,
authenticity, feeling distinctive, psychological ownership, p = .00003) between the no AI hint and AI hint conditions
and credibility showed no significant differences, with for both product forms, but the effect was more
small effect sizes. This suggests that while aspects like pronounced for canvas prints. Quality was rated lower for
cost, risk perception, and quality are significantly canvas prints with an AI hint compared to those without
influenced by the product form, other factors remain (diff = -1.496, p < .0001, M_noAI_canvas=3.10 vs
relatively unaffected by whether content is an e-book or M_AI_canvas=3.07). In books, the presence of an AI hint also
print canvas. resulted in lower quality ratings, demonstrating the
pervasive impact of AI involvement across different
Interactions. For the perception of financial risk, a
mediums (diff = -1.167, p < .0001, M_noAI_Book=4.57 vs
significant interaction effect was identified between the
M_AI_Book=3.40). Product attitudes were significantly
product form (book vs. canvas print) and AI involvement
influenced (F(1,294) = 4.581, p = .0332) by the interaction
(no AI hint vs. AI hint) (F(1,294) = 3.94, p = .048), Table 2.
between product form and AI involvement. For books, the
This interaction indicates that the combined influence of
presence of an AI hint notably decreased positive attitudes
product form and AI hint alters consumer perceptions of
towards the product (diff = -1.083, p = .0002,
financial risk. The Tukey post-hoc comparisons revealed
M_noAI_Book=4.67 vs M_AI_Book=3.59). Conversely, for
specific differences across the groups: the financial risk
canvas prints, while there was still a decline in positive
was perceived to be higher for books with an AI hint
attitudes with an AI hint, the difference was less
compared to books without an AI hint (diff = 0.82, p =
pronounced and not statistically significant, suggesting that
.019, M_noAI_Book=1.96 vs M_AI_Book=2.78). Conversely,
the impact of AI hints might be more critical in content-
there was no significant difference in financial risk
based formats (diff = -0.303, p = .6427, M_noAI_canvas=4.46
perception between canvas prints without an AI hint and
vs M_AI_canvas=4.16). The ease of copying showed a
those with an AI hint (diff = 0.036, p = .999,
significant interaction effect (F(1,294) = 8.360, p = .0041)
M_noAI_canvas=2.31 vs M_AI_canvas=2.35).
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1535
Table 2. Significant interactions between different product forms and AI hints
Mean_no SD_noAI_B Mean_A SD_AI_Bo Mean_noA SD_noAI_c Mean_A SD_AI_canva p-
Variable F value
AI_Book ook I_Book ok I_canvas anvas I_canvas s value
Financialrisk 1.96 1.41 2.78 1.94 2.31 1.73 2.35 1.68 3.939 0.0481
Enjoyment 4.31 1.58 3.62 1.88 4.36 1.69 4.52 1.65 4.508 0.0346
Permanence 4.58 1.55 4.13 1.75 4.59 1.64 3.35 1.77 4.058 0.0449
Credible 4.75 1.16 3.40 1.81 4.08 1.46 3.88 1.63 10.508 0.0013
0.0000
Quality 4.57 1.39 3.40 1.57 3.10 0.52 3.07 0.83 18.072
3
product_attitu
4.67 1.35 3.59 1.53 4.46 1.64 4.16 1.73 4.581 0.0332
des
Easytocopy 4.04 1.45 4.83 1.34 4.91 1.46 4.71 1.66 8.360 0.0041
Mediation effects. Performance Risk served as a mediator 0.306, CI [-0.540, -0.080] (Table 4).
in several relationships - it mediated the impact of AI
Perceived Authenticity significantly mediated the
involvement on WTP, decreasing willingness to pay with a
relationship between AI involvement and several
mediation effect of -2.944, CI [-5.325, -0.691].
outcomes: it negatively influenced WTP, with a mediation
Performance risk also influenced willingness to buy
effect of -3.587, CI [-5.732, -1.729], perceived authenticity
(WTB), with a negative mediation effect of -0.347, CI [-
also decreased WTB, with a mediation effect of -0.523, CI
0.628, -0.093]. Additionally, performance risk mediated
[-0.808, -0.259] and product attitudes, the mediation effect
the impact on product attitudes, showing an effect of -
was -0.478, CI [-0.718, -0.240].
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1536
Table 4. Mediation effects of perceived performance risk and authenticity
Response CI CI
Cond Mediator Estimate
Variable Lower Upper
Ai
Performance Risk willingness to pay -2.944 -5.325 -0.691
involvement
Ai
Performance Risk willingness to buy -0.347 -0.628 -0.093
involvement
Ai
Performance Risk product attitudes -0.306 -0.540 -0.080
involvement
Ai Perceived
willingness to pay -3.587 -5.732 -1.729
involvement Authenticity
Ai Perceived
willingness to buy -0.523 -0.808 -0.259
involvement Authenticity
Ai Perceived
product attitudes -0.477 -0.718 -0.240
involvement Authenticity
Study 1 results discussion. Key findings reveal that AI Performance risk and perceived authenticity were crucial
involvement hints significantly affect consumer mediators in the relationship between AI involvement and
perceptions. Specifically, the AI hints led to lower consumer responses, with increased performance risk
willingness to pay and buy, authenticity, indicating associated with AI-authored products lowering financial
consumer reservations about AI-produced content, commitment and purchase intent. Perceived authenticity
possibly due to concerns about authenticity or the novelty significantly decreased willingness to pay, willingness to
of AI in creative production (Wu et al., 2024). buy, and product attitudes, highlighting the importance of
Furthermore, the presence of AI hints was associated with enhancing the perceived genuineness of AI-generated
increased perceived financial and performance risks, which products to appeal to potential buyers (Tsai et al., 2024).
might reflect consumer uncertainty about the reliability and H7, H8 confirmed.
value retention of AI-generated products (Deryl et al.,
3.2. Study 2: Evaluating the Effectiveness of Enhanced
2023). H1, H2, H3 confirmed
Product Information on Mitigating Perceived Risks
The interaction effect on financial risk perceptions of AI-Created Products
indicated a higher perceived financial risk for books when
Main effects of AI cues. For participants in the condition
AI involvement was hinted, suggesting the deep content-
where the author was specified as human versus when it
based format may enhance uncertainty about product
was indicated that the author is AI in context of e-books,
quality. In contrast, AI hints did not significantly alter risk
the analysis of variance (ANOVA) revealed significant
perceptions for canvas prints, potentially due to the visible
differences across multiple variables (Table 4). Perceived
outcome of these art-based products providing a sense of
authenticity varied significantly between conditions, with
security. Additionally, AI involvement negatively
participants perceiving higher authenticity when the author
impacted quality perceptions and attitudes towards AI-
was human (M_human=5.172, SD=1.352 vs. M_AI=3.806,
involved books, underscoring challenges AI faces in being
SD=1.872, F(1,124)=21.92, p<.00001, η²=.17). Similarly,
accepted in creative processes (Horton Jr et al., 2023). H1
enjoyment levels were notably different, favouring the
confirmed again.
human-authored condition (M_human=5.219, SD=1.419 vs.
The study's findings emphasize the significant role of trust M_AI=4.145, SD=1.587, F(1,124)=15.98, p=.00011, η²=.12),
in artificial intelligence in moderating consumer behaviors the quality of the ebook was perceived as higher when
towards AI-involved products. Trust in AI increased authored by a human (M_human=4.906, SD=1.411 vs.
purchasing likelihood when AI involvement was disclosed M_AI=4.226, SD=1.509, F(1,124)=6.83, p=.010, η²=.05).
and mitigated perceived performance risks, improving
In addition, financial and performance risks were
attitudes towards AI-generated products (Lukyanenko et
perceived as significantly lower when the author was
al., 2022). Experience with AI-generated content also
human (financial risk: M_human=1.859, SD=1.39 vs.
played a moderating role, leading to lower perceived
M_AI=2.484, SD=1.844, F(1,124)=4.59, p=.034, η²=.04;
financial risks and more positive attitudes towards AI
performance risk: M_human=3.359, SD=1.505 vs.
products (Chen et al., 2022). H4, H5 confirmed.
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1537
M_AI=4.081, SD=1.768, F(1,124)=6.06, p=.015, η²=.05). M_AI=3.177, SD=2.045, F(1,124)= 7.99, p=.0055, η²=.06),
Lastly, participants showed a higher willingness to buy but willingness to pay was insignificant.
when the author was human (M_human=4.172, SD=1.899 vs.
Table 5. Consumer perceptions of the e-book based on author (author was human vs AI)
Variable Mean_human SD_human Mean_AI SD_AI F_statistic p_value η2
Willingness to pay 9.607 6.059 9.447 9.85 0.01 0.9132986 0
Willingness to buy 4.172 1.899 3.177 2.045 7.99 0.005492656 0.06
Financial risk 1.859 1.39 2.484 1.844 4.59 0.03438804 0.04
Performance risk 3.359 1.505 4.081 1.768 6.06 0.01522413 0.05
Authentic 5.172 1.352 3.806 1.872 21.92 8.100045e-06 0.17
Enjoyment 5.219 1.419 4.145 1.587 15.98 0.0001101996 0.12
Quality 4.906 1.411 4.226 1.509 6.83 0.01010112 0.05
Low and high information main effects. The analysis of the perceived less risk in the high information condition
main effects between conditions with low and high (M_high=3.431, SD=1.639 vs. M_low=4.016, SD=1.668;
information availability provided insights into consumer t=1.99, F(1,124)=3.94, p=0.049, η²=0.03). The participants in
responses across several metrics. While most variable the high information condition exhibited slightly higher
showed statistically significant differences after correcting willingness to pay and buy, along with lower perceptions
for multiple comparisons, trends suggest possible of financial and performance risks compared to those in the
influences of information level on consumer perceptions low information condition, but differences were not
and intentions: marginally significant difference was significant (Table 6).
observed for performance risk, where participants
Table 6: Consumer perceptions based on information cues (low and high information availability)
Variable Mean_low SD_cond_0 Mean_high SD_cond_1 F_statistic p_value η2
Willingness to pay 8.759 7.83 10.25 8.37 1.07 0.3032581 0.01
Willingness to buy 3.426 1.97 3.923 2.064 1.91 0.1693285 0.02
Financial risk 2.295 1.657 2.046 1.653 0.71 0.4004156 0.01
Performance risk 4.016 1.668 3.431 1.639 3.94 0.0492581 0.03
Authentic 4.311 1.794 4.677 1.724 1.36 0.2464671 0.01
Enjoyment 4.508 1.49 4.862 1.676 1.57 0.2128384 0.01
Quality 4.344 1.425 4.785 1.536 2.79 0.09752671 0.02
Interactions. The analysis of the interaction effects with increased trust enhancing the likelihood of purchasing
between the authorship (human vs. AI) and AI AI-authored products (WTB: β = 0.461, p = 0.01130;
involvement hints across various consumer perception WTP: β = 2.271, p = 0.00232). These findings suggest that
variables did not yield any statistically significant results in trust in AI not only reduces skepticism but also boosts
context of e-books. This indicates that risk perceptions are consumer engagement and financial commitment to
primarily based on AI involvement and are hardly affected products associated with AI (Table 7). Additionally,
by presenting more information. familiarity with AI-generated content played a crucial role
in moderating perceptions of financial risk associated with
Moderators. Trust in AI significantly influenced
AI-authored products. Consumers with more frequent
perceptions of authenticity, with greater trust leading to
interactions with AI content exhibited lower perceived
higher perceived authenticity in products identified as AI-
financial risks (β = 0.371, p = 0.01634), indicating that
authored (β = 0.398, p = 0.00755). Similarly, trust in AI
familiarity with AI can mitigate concerns and enhance
positively moderated consumer willingness to buy and pay,
comfort with AI-driven products (Table 7).
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1538
Table 7. Moderation effects of AI involvement (vs human) and consumer behavior
Moderator Response Variable Coefficient p-value
Trust in artificial intelligence Authenticity 0.3981564 0.00755
Familiarity with AI-generated content Financial Risk 0.3710492 0.01634
Trust in artificial intelligence Willingness to Buy 0.4610359 0.01130
Trust in artificial intelligence Willingness to Pay 2.2713829 0.00232
Mediations. Performance risk and perceived authenticity, significantly dampen consumer willingness to invest,
as shown in the Table 8, served as significant mediators in willingness to buy and quality perceptions of AI-authored
the relationship between the type of authorship (human vs. products. Perceived authenticity significantly mediated the
AI) and various consumer responses in the experiment. relationship between authorship type and willingness to
Performance Risk negatively mediated the impact of AI pay, with a substantial negative effect (-3.520, CI [-5.545, -
authorship on willingness to pay (-1.451, CI [-2.879, - 1.851]), willingness to buy (-1.176, CI [-1.720, -0.651]),
0.301]), willingness to buy (-0.467 (CI [-0.877, -0.099]), quality perception (-0.439, CI [-0.746, -0.192]) reinforcing
perceived product quality (-0.249, CI [-0.494, -0.045]). the importance of authenticity in consumer purchase
This suggests that perceptions of performance risk decisions (Table 8).
Table 8. Mediation effects of perceived performance risk, authenticity for human vs AI involvement and consumer
behavior.
Condition Mediator Response Variable Estimate CI Lower CI Upper
AI involvement Performance Risk Willingness to Pay -1.451 -2.879 -0.301
AI involvement Performance Risk Willingness to Buy -0.467 -0.878 -0.099
AI involvement Performance Risk Quality -0.249 -0.494 -0.045
AI involvement Perceived Authenticity Willingness to Pay -3.520 -5.545 -1.851
AI involvement Perceived Authenticity Willingness to Buy -1.176 -1.720 -0.651
AI involvement Perceived Authenticity Quality -0.439 -0.746 -0.192
Study 2 results discussion. Building on findings from cues (Lefkeli et al., 2024). H6 rejected.
Study 1, which indicated detrimental effects of AI
Moderation analyses showed that trust in AI and
involvement cues on product attitudes and consumer
experience with AI-generated content significantly
behavior, Study 2 aimed to evaluate whether providing
influenced consumer responses. Higher trust in AI was
additional content details, such as detailed summaries,
associated with enhanced perceptions of authenticity and
tables of contents, and sample chapters, could mitigate
increased willingness to engage financially with AI-created
perceived risks associated with AI-created products.
products. Conversely, consumers with more frequent
However, key findings revealed significant main effects
interactions with AI content exhibited reduced financial
where human-authored products were consistently
risk perceptions, suggesting that familiarity might alleviate
perceived as more authentic, enjoyable, and of higher
apprehension. H4, H5 confirmed again.
quality than those attributed to AI. These products evoked
lower financial and performance risks and were more Mediation analyses further highlighted that performance
likely to be purchased by participants. H1,H2,H3 risk and perceived authenticity are crucial in mediating the
confirmed again. impact of AI authorship on various consumer responses.
Notably, both performance risk and authenticity
The introduction of detailed information, regardless of the
perceptions negatively mediated willingness to pay, buy,
amount (high vs. low), did not significantly alter
and perceived quality, underscoring their key roles in the
willingness to pay, buy, product perceptions, or
acceptance and valuation of AI-generated content. H7, H8
significantly change perceived risks—indicating that
confirmed again.
adding more product information may not be sufficient to
alter established perceptions created by AI involvement
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1539
3.3. Study 3: "The Role of Outcome Visibility in as more authentic. Additionally, the perceived effort put
Modulating Consumer Reactions to AI and Artist into the canvas creation was viewed as significantly higher
Authored Products" in the artist-authored condition (M_artist=4.000, SD=1.246
vs. M_AI=3.214, SD=1.480, F(1,200)=16.55, p<.0001,
Main effects of AI involvement. For participants in the
η²=.08), suggesting that participants value human
condition where the author was specified as an artist versus
involvement as indicative of greater effort. However, other
when it was indicated that the author is AI in context of
variables such as willingness to pay, willingness to buy,
canvas prints, the analysis of variance (ANOVA) revealed
financial risk, performance risk, enjoyment, and quality
significant differences across multiple variables. Notably,
showed no significant differences, indicating that these
perceived authenticity was significantly higher when the
perceptions were not substantially impacted by whether the
author was specified as an artist (M_artist=4.452, SD=1.427
author was artist or AI, suggesting that perception of art-
vs. M_AI=3.765, SD=1.661, F(1,200)=9.88, p=.0019, η²=.05),
based products, like canvas prints are less affected by AI
indicating that participants perceived artist-authored works
involvement (Table 9).
Table 9. Perceptions of Authenticity and Effort in Artist vs. AI-Authored Works
Variable Mean_artist Mean_AI SD_artist SD_AI F_statistic p-value R2
Willingness to pay 22.845 19.907 13.306 12.132 2.68 0.1031769 0.01
Willingness to buy 3.01 2.949 1.908 1.858 0.05 0.8192418 0
Financial risk 2.846 3.071 1.904 1.885 0.71 0.399246 0
Performance risk 3.962 4.041 1.848 1.942 0.09 0.7669051 0
Authentic 4.452 3.765 1.427 1.661 9.88 0.001939864 0.05
Enjoyment 4.5 4.367 1.461 1.522 0.4 0.5287127 0
Quality 4.202 4.163 1.242 1.306 0.05 0.8297026 0
6.923171e-
Effort 4 3.214 1.246 1.48 16.55 0.08
05
For participants in the condition where the outcome of the perceived authenticity, enjoyment, or overall quality of the
canvas print was blurred versus when the outcome was canvas print between the blurred and visible conditions,
fully visible, the analysis of variance (ANOVA) revealed suggesting that these aspects of consumer perception might
significant differences across several key variables. be less influenced by the visibility of the outcome or
Willingness to pay (M_blurred=18.475, SD=12.608 vs. possibly overshadowed by more immediate concerns
M_visible=24.47, SD=12.347, F(1,200)=11.6, p=.0008, η²=.06) related to financial and performance risks associated with
and willingness to buy (M_blurred=2.641, SD=1.703 vs. the purchase decision.
M_visible=3.333, SD=1.995, F(1,200)=7.02, p=.009, η²=.04)
Interactions. The analysis of the interaction effects
was significantly higher when the outcome was visible,
between the authorship (human vs. AI) and the clarity of
indicating that seeing the final product can substantially
the outcome (blurred vs. visible) revealed significant
enhance higher purchase intent when consumers can fully
interaction influences on several key consumer
assess the product. In terms of perceived risks, financial
perceptions. Notably, there was a significant interaction
risk perceptions were significantly lower when the
effect on perceived effort, suggesting that the perception of
outcome was visible (M_blurred=3.757, SD=1.85 vs.
effort exerted in creating the canvas print was influenced
M_visible=2.121, SD=1.554, F(1,200)=46.48, p<.00001,
by both the type of authorship and the visibility of the
η²=.19), and performance risk followed a similar trend
outcome (F(1,197) = 16.99, p<.0001, η²=.08). The Tukey
(M_blurred=5.000, SD=1.793 vs. M_visible=2.960, SD=1.355,
post-hoc comparisons revealed that the effort was
F(1,200)=83.67, p<.00001, η²=.31), indicating that
perceived as significantly lower when the author was AI
transparency about the product outcome reduces
and the outcome was visible compared to when the author
uncertainty and enhances confidence in the product's
was human and the outcome was blurred, underscoring
quality and worth.
how AI authorship negatively impacts perceived effort
However, there were no significant differences in especially when the product outcome is clear (Table 10).
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1540
Table 10. Interaction effects of AI involvement (vs artist) and level of outcome visibility
Variable Mean Sq F value Pr(>F)
Willingness to pay 764.194 5.0536 0.0257
Willingness to buy 0.5 0.1445 0.7043
Financial risk 1.048 0.3566 0.5511
Performance risk 0.081 0.0315 0.8593
Authentic 31.594 14.0711 0.0002
Enjoyment 0.96 0.4289 0.5133
Quality 1.058 0.6508 0.4208
Effort 27.593 16.9870 0.0001
Furthermore, the interaction effects were significant for there were no significant differences in willingness to pay,
willingness to pay (F(1,197) = 5.05, p=.0257, η²=.03), where buy, and risk perceptions based on authorship type alone
participants were willing to pay more when the outcome for art-based products, suggesting that these factors might
was visible, particularly when the author was human. be less susceptible to changes based solely on who created
(Table 10). This indicates that visibility of the final product the art.
combined with artist authorship favorably influences
The visibility of the final product had a significant impact
consumer valuation of the artwork. Authenticity also
on willingness to pay and buy, with higher ratings in the
showed significant interaction effects (F(1,197) = 14.07,
visible condition. This suggests that transparency in the
p<.0001, η²=.05), with authenticity perceived higher in
final product boosts consumer confidence and willingness
artist-authored works, especially when the outcome was
to invest. Financial and performance risk perceptions were
visible, suggesting that visibility amplifies the perceived
significantly lower when the outcome was visible,
authenticity when the creator is known to be human
emphasizing the importance of product transparency in
(artist).
reducing consumer uncertainty (Sansome et al., 2024;
Mediations. Perceived authenticity served as a significant Busser & Shulga, 2019).
mediator in the relationship between the type of authorship
Significant interaction effects were found on perceived
(artist vs. AI) and consumer financial decisions regarding
effort, willingness to pay, and authenticity, indicating that
canvas prints. It significantly negatively mediated the
these perceptions are jointly influenced by authorship type
impact of AI authorship on willingness to pay (-2.097, CI
and product visibility. Notably, perceived effort was lowest
[-3.769, -0.773]) and willingness to buy (-0.350, CI [-
when AI was the author and the outcome was visible,
0.613, -0.123]). This indicates that the authenticity
possibly due to assumptions that automated processes
attributed to the product by the author type directly affects
require less human effort. However, there were no
purchase intent and level of investment in these products.
interaction effects related to perceived risks, suggesting
Discussion of Study 3 results. Study 3 explored the that risk perceptions might be primarily influenced by AI
influence of the visibility of final product outcomes on cues and are less affected by additional information and
consumer perceptions, specifically when authorship was outcome visibility. H9 rejected.
identified as either an artist or AI, building on insights
Perceived authenticity mediated the relationship between
from Study 2 that suggested additional product information
authorship type and consumer financial decisions,
might not significantly alter perceptions of AI-generated
significantly affecting willingness to pay and buy. This
products.
mediation underscores the critical role of authenticity in
Significant differences were observed in perceived consumer valuation, especially in contexts where
authenticity and effort depending on the authorship technology replaces traditional creative roles (Nunes et al.,
condition. Participants rated products as more authentic 2021). H8 confirmed again.
and indicative of more effort when the author was
In conclusion, Study 3 reveals that while the visibility of
identified as an artist compared to an AI, highlighting a
the final product can enhance consumer perceptions and
preference for human involvement in creative processes,
mitigate risk concerns, the nature of authorship
which are perceived as more genuine and labor-intensive
significantly influences how authenticity and effort are
(Horton et al., 2023). H2 confirmed again. However,
perceived. However, only perceived authenticity is
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1541
significantly affected in the mix of these two conditions, commitment and purchase intent. This underlines the
later mediating financial decisions. importance of addressing these perceptual factors in
marketing strategies for AI-generated products (Naz &
4. Overall Results
Kashif, 2024).
The series of studies presented aimed to explore how AI
Interactions and Visibility of Product Outcomes. Study 3's
involvement and various levels of product information
exploration into the visibility of the final product outcome
affect consumer perceptions and behaviors across different
revealed that clear visibility could significantly reduce
product forms, specifically books, e-books and canvas
perceived risks and enhance consumer confidence,
prints. A consistent theme across all studies was the impact
especially when combined with human authorship. This
of AI involvement cues on consumer attitudes,
was particularly notable in the canvas print context, where
underscoring significant concerns regarding authenticity,
visibility of the final product helped alleviate concerns
credibility, and the perceived effort of AI-generated
over financial and performance risks, but did not
products compared to those attributed to human creators.
counteract the negative biases induced by AI involvement
Impact of AI Involvement. AI involvement hints (Abràmoff et al., 2023).
consistently led to negative perceptions and willingness to
Overall, these studies highlight a complex interplay
buy and pay, particularly in terms of authenticity and
between AI involvement, product transparency, consumer
credibility. Consumers showed reservations about AI-
trust, and perceived authenticity. They illustrate that while
produced content, likely stemming from concerns over the
AI involvement can detrimentally affect consumer
novelty of AI in creative production and its potential
perceptions, strategies that enhance trust in AI, improve
implications for quality and reliability. This was evident in
familiarity with AI products, and ensure transparency in
both content based products (books, e-books) and art-
product outcomes can potentially mitigate these effects.
related products (canvas prints), although the art based
The findings suggest that increasing consumer exposure to
nature of canvas prints seemed to mitigate some of the
AI and enhancing the perceived genuineness of AI-
negative perceptions associated with AI involvement,
generated products are vital for fostering acceptance and
suggesting that art-based product forms might offer a sense
integrating AI into creative industries effectively.
of security that deep-content formats lack (Neyazi et al.,
2023) 5. Discussion
Role of Trust and Familiarity. Across the studies, trust in Theoretical Implications. The study explored consumer
AI and familiarity with AI-generated content emerged as perceptions of products based on authorship (human, artist,
significant moderators of consumer responses. Higher AI) and type (books, e-books, print canvas), enhancing
levels of trust in AI technology and more frequent theoretical insights into consumer psychology amid AI
interactions with AI-generated content were linked to more advancements (Mariani et al., 2022). Interaction effects
positive attitudes toward AI-involved products. This underscore the need to consider variables like product type
suggests that enhancing consumer trust and familiarity and AI involvement. The study identifies performance risk
with AI could play a crucial role in improving market and perceived authenticity as key mediators affecting
acceptance and success of AI-driven innovations (Choung consumer decisions (Hirunyawipada & Paswan, 2006;
et al., 2023; Frank et al., 2023). Nunes et al., 2021; Safeer et al., 2023). Additionally, trust
in AI and familiarity with AI-generated content emerged as
Effectiveness of Additional Product Information. Study 2
significant moderators, highlighting their influence on
specifically investigated whether providing detailed
consumer responses and attitudes (Lukyanenko et al.,
product information could alter the negative perceptions
2022). It expands also AI disclosure literature based on
elicited by AI involvement hints. The findings indicated
brands (Lefkeli et al., 2024).
that while additional information did not significantly
change willingness to pay or buy, it did not substantially Practical Implications. The study's practical implications
alter perceived risks. This suggests that while detailed offer valuable insights for product development,
product information is beneficial, it may not be sufficient marketing, and risk management. Companies can tailor
on its own to counteract the negative biases induced by AI products to emphasize human involvement in AI generated
involvement cues (Craig & Choi, 2024). products, enhancing perceived value and authenticity
(Nunes et al., 2021). Understanding consumer concerns
Mediation by Authenticity and Performance Risk. The
about AI-related risks enables firms to develop mitigation
mediation analyses across the studies highlighted that
strategies effectively (Kalogiannidis et al., 2024).
perceived authenticity and performance risk significantly
Transparency in AI involvement, but highlighting human
influenced consumer behaviors and attitudes. Increased
added value in creation process can foster trust and
perceptions of risk and decreased authenticity associated
confidence (Zerilli et al., 2022). Personalized marketing
with AI-authored products were found to lower financial
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1542
strategies, based on consumer trust and familiarity with AI, performance reliability. However, these negative effects
can improve engagement. Educational initiatives about AI can be moderated by the consumers' trust in AI and their
capabilities can boost trust and acceptance (Spais et al., familiarity with AI-generated content. When consumers
2023) Lastly, the study emphasizes the need for ethical trust AI technology and are familiar with AI products, their
considerations in AI integration, promoting transparency perceptions tend to be less negative, and their willingness
and authenticity in creative processes (Nguyen et al., to engage with these products increases.
2023).
Furthermore, the study found that providing detailed
Limitations of this study. Limitations of the study include product information can alter perceived risks associated
potential sample bias due to limited diversity in with AI-generated products. This suggests that
demographics, affecting the findings' generalizability. transparency about the AI's role and the functionalities of
Focusing on specific product types (books. e-books, print the products can help mitigate some of the consumers'
canvas) limits the study's applicability to other product concerns.
categories. The study didn't differentiate between various
In terms of practical implications, the study advises
AI capabilities or account for potential ethical concerns in
businesses to carefully consider how they disclose AI
AI involvement. The cross-sectional design captures a
involvement in product creation. To counteract potential
single snapshot of consumer attitudes, potentially missing
negativity, businesses should focus on building consumer
temporal changes. Additionally, unexplored external
trust and increasing familiarity with AI-generated products
factors like marketing strategies, peer influence, and media
through educational initiatives and transparent marketing
portrayal could influence consumer perceptions, and
strategies. Additionally, emphasizing the human oversight
reliance on self-reported measures may introduce bias.
and ethical standards adhered to in the AI's integration into
Future research directions. Future studies should use product development may further alleviate consumer
cross-cultural and longitudinal methods to examine how apprehensions.
consumers' opinions of AI products are changing.
These findings contribute to the academic literature by
Examining certain AI traits like intellect and creativity
providing a detailed analysis of how AI involvement
might uncover important variables influencing consumer
impacts consumer behavior through the lenses of perceived
choices. To increase trust and adoption rates, consideration
risk and authenticity. They also explore how the negative
should be given to the ethical implications of AI's creative
impacts of AI disclosures can be mitigated by enhancing
process as well as the efficacy of consumer education
consumer trust and familiarity with AI technologies.
initiatives. It's critical to do research on AI's involvement
in real product personalization. Future research must also Overall, the study highlights the complexities of
focus on comprehending AI's effects on various industries, integrating AI into consumer product markets and
using experimental methods to get causal insights, and underscores the need for strategic transparency and
evaluating how technology developments affect customer consumer education to foster a more accepting consumer
preferences and market dynamics. environment for AI-generated products.
6. Conclusions References:
The research investigated the consequences of disclosing [1] Abràmoff, M. D., Tarver, M. E., Loyo-Berrios, N.,
AI involvement in product creation on consumer Trujillo, S., Char, D., Obermeyer, Z., Eydelman, M.
perceptions, specifically focusing on authenticity, B., Foundational Principles of Ophthalmic Imaging
perceived risks, and the moderating roles of trust in AI and and Algorithmic Interpretation Working Group of the
familiarity with AI-generated content. The results clearly Collaborative Community for Ophthalmic Imaging
demonstrated that disclosing AI involvement generally Foundation, Washington, D.C., & Maisel, W. H.
leads to negative consumer perceptions, particularly in (2023). Considerations for addressing bias in artificial
terms of reduced authenticity and increased perceived risks intelligence for health equity. Npj Digital Medicine,
for AI generated products. These factors substantially 6(1), 170. https://doi.org/
lower consumer willingness to engage financially with AI- 10.1038/s41746-023-00913-9
generated products. [2] Agarwal, S., & Teas, R. K. (2001). Perceived Value:
Perceived authenticity and risks were identified as Mediating Role of Perceived Risk. Journal of
significant mediators in the relationship between AI Marketing Theory and Practice, 9(4), 1–14.
involvement and the diminished consumer willingness to https://doi.org/10.1080/10696679.2001.11501899
purchase. The data suggest that consumers are wary of [3] Aldousari, A. A., Yasmin, F., Yajid, M. S. A., &
products when AI's role in their creation is highlighted, Ahmed, Z. U. (2017). Consumer evaluations of store
doubting both the genuineness of the products and their brands: Effects of product-perceived risks. Journal of
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1543
Transnational Management, 22(2), 71–90. K.-U., & Von Wangenheim, F. (2020). Artificial
https://doi.org/10.1080/15475778.2017.1302782 intelligence and marketing: Pitfalls and opportunities.
Journal of Interactive Marketing, 51(1), 91–105.
[4] Atasoy, O., & Morewedge, C. K. (2018). Digital
goods are valued less than physical goods. Journal of [15] de Kerviler, G., Heuvinck, N., & Gentina, E. (2021).
Consumer Research, 44(6), 1343–1357. “Make an Effort and Show Me the Love!” Effects of
Indexical and Iconic Authenticity on Perceived Brand
[5] Bauer, R. A. (1960). Consumer behavior as risk
Ethicality. Journal of Business Ethics.
taking. Proceedings of the 43rd National Conference
https://doi.org/10.1007/s10551-021-04779-3
of the American Marketing Association.
[16] Deryl, M. Ds., Verma, S., & Srivastava, V. (2023).
[6] Bellaiche, L., Shahi, R., Turpin, M. H.,
How does AI drive branding? Towards an integrated
Ragnhildstveit, A., Sprockett, S., Barr, N.,
theoretical framework for AI-driven branding.
Christensen, A., & Seli, P. (2023). Humans versus
International Journal of Information Management
AI: Whether and why we prefer human-created
Data Insights, 3(2), 100205. https://doi.org/10.1016/j.
compared to AI-created artwork. Cognitive Research:
jjimei.2023.100205
Principles and Implications, 8(1), 42.
https://doi.org/10.1186/s41235-023-00499-6 [17] Dhagarra, D., Goswami, M., & Kumar, G. (2020).
Impact of Trust and Privacy Concerns on Technology
[7] Betzing, J. H., Tietz, M., Vom Brocke, J., & Becker,
Acceptance in Healthcare: An Indian Perspective.
J. (2020). The impact of transparency on mobile
International Journal of Medical Informatics, 141,
privacy decision making. Electronic Markets, 30(3),
104164. https://doi.org/10.1016/j.ijmedinf.2020.
607–625. https://doi.org/10.1007/s125
104164
25-019-00332-3
[18] Frank, D.-A., Jacobsen, L. F., Søndergaard, H. A., &
[8] Biswas, D., & Biswas, A. (2004). The diagnostic role
Otterbring, T. (2023). In companies we trust:
of signals in the context of perceived risks in online
Consumer adoption of artificial intelligence services
shopping: Do signals matter more on the Web?
and the role of trust in companies and AI autonomy.
Journal of Interactive Marketing, 18(3), 30–45.
Information Technology & People, 36(8), 155–173.
https://doi.org/10.1002/dir.20010
https://doi.org/10.
[9] Brendel, A. B., Mirbabaie, M., Lembcke, T.-B., & 1108/ITP-09-2022-0721
Hofeditz, L. (2021). Ethical management of artificial
[19] Gupta, S., Kushwaha, Pooja. S., Badhera, U.,
intelligence. Sustainability, 13(4), 1974.
Chatterjee, P., & Gonzalez, E. D. R. S. (2023).
[10] Cetinic, E., & She, J. (2021). Understanding and Identification of benefits, challenges, and pathways in
Creating Art with AI: Review and Outlook E-commerce industries: An integrated two-phase
(arXiv:2102.09109). arXiv. http://arxiv. decision-making model. Sustainable Operations and
org/abs/2102.09109 Computers, 4, 200–218.
[11] Chen, Y., Prentice, C., Weaven, S., & Hisao, A. https://doi.org/10.1016/j.susoc.2023.08.005
(2022). The influence of customer trust and artificial [20] Ha, H.-Y. (2006). The Effects of Consumer Risk
intelligence on customer engagement and loyalty – Perception on Pre-purchase Information in Online
The case of the home-sharing industry. Frontiers in Auctions: Brand, Word-of-Mouth, and Customized
Psychology, 13, 912339. https://doi.org/10.3389/ Information. Journal of Computer-Mediated
fpsyg.2022.912339 Communication, 8(1), 0–0.
[12] Choung, H., David, P., & Ross, A. (2023). Trust in https://doi.org/10.1111/j.10
AI and Its Role in the Acceptance of AI 83-6101.2002.tb00160.x
Technologies. International Journal of Human– [21] Haleem, A., Javaid, M., Asim Qadri, M., Pratap
Computer Interaction, 39(9), 1727–1739. Singh, R., & Suman, R. (2022). Artificial intelligence
https://doi.org/ (AI) applications for marketing: A literature-based
10.1080/10447318.2022.2050543 study. International Journal of Intelligent Networks,
[13] Craig, M. J. A., & Choi, M. (2024). The role of 3, 119–132. https://doi.org/10.1016/j.ijin.
affective and cognitive involvement in the mitigating 2022.08.005
effects of AI source cues on hostile media bias. [22] Hirunyawipada, T., & Paswan, A. K. (2006).
Telematics and Informatics, 88, 102097. Consumer innovativeness and perceived risk:
https://doi.org/10.1016/j.tele.2024.102097 Implications for high technology product adoption.
[14] De Bruyn, A., Viswanathan, V., Beh, Y. S., Brock, J. Journal of Consumer Marketing, 23(4), 182–198.
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1544
https://doi.org/10.1108/07363760610674310 [33] Luce, L. (2018). Artificial intelligence for fashion:
How AI is revolutionizing the fashion industry.
[23] Hong, J.-W. (2018). Bias in perception of art
Apress.
produced by artificial intelligence. 290–303.
[34] Lukyanenko, R., Maass, W., & Storey, V. C. (2022).
[24] Horowitz, M. C., Kahn, L., Macdonald, J., &
Trust in artificial intelligence: From a Foundational
Schneider, J. (2023). Adopting AI: How familiarity
Trust Framework to emerging research opportunities.
breeds both trust and contempt. AI & SOCIETY.
Electronic Markets, 32(4), 1993–2020.
https://doi.org/10.1007/s00146-023-01666-5
https://doi.org/10.1007/s12525-022-00605-4
[25] Horton Jr, C. B., White, M. W., & Iyengar, S. S.
[35] Mariani, M. M., Perez‐Vega, R., & Wirtz, J. (2022).
(2023). Bias against AI art can enhance perceptions
AI in marketing, consumer research and psychology:
of human creativity. Scientific Reports, 13(1), 19001.
A systematic literature review and research agenda.
https://doi.org/10.1038/s41598-023-45202-3
Psychology & Marketing, 39(4), 755–776.
[26] Ismatullaev, U. V. U., & Kim, S.-H. (2024). Review
[36] Marti, C. L., Liu, H., Kour, G., Bilgihan, A., & Xu,
of the factors affecting acceptance of AI-infused
Y. (2024). Leveraging artificial intelligence in firm-
systems. Human Factors, 66(1), 126–144.
generated online customer communities: A
[27] Kalogiannidis, S., Kalfas, D., Papaevangelou, O., framework and future research agenda. Journal of
Giannarakis, G., & Chatzitheodoridis, F. (2024). The Service Management.
Role of Artificial Intelligence Technology in
[37] Naz, H., & Kashif, M. (2024). Artificial intelligence
Predictive Risk Assessment for Business Continuity:
and predictive marketing: An ethical framework from
A Case Study of Greece. Risks, 12(2), 19.
managers’ perspective. Spanish Journal of Marketing
https://doi.org/10.3390/risks12020019
- ESIC. https://doi.org/10.
[28] Kang, J., & Kim, S.-H. (2013). What Are Consumers 1108/SJME-06-2023-0154
Afraid of? Understanding Perceived Risk toward the
[38] Newman, G. E., & Bloom, P. (2012). Art and
Consumption of Environmentally Sustainable
authenticity: The importance of originals in
Apparel. Family and Consumer Sciences Research
judgments of value. Journal of Experimental
Journal, 41(3), 267–283. https://doi.org/10.
Psychology: General, 141(3), 558.
1111/fcsr.12013
[39] Neyazi, T. A., Ng, S. W. T., Hobbs, M., & Yue, A.
[29] Kelly, S., Kaye, S.-A., & Oviedo-Trespalacios, O.
(2023). Understanding user interactions and
(2023). What factors contribute to the acceptance of
perceptions of AI risk in Singapore. Big Data &
artificial intelligence? A systematic review.
Society, 10(2), 20539517231213823.
Telematics and Informatics, 77, 101925.
https://doi.org/10.1177/20539517231213823
https://doi.org/10.1016/j.tele.2022.101925
[40] Nguyen, A., Ngo, H. N., Hong, Y., Dang, B., &
[30] Kim, N., & Jin, B. E. (2018). Why Buy New When
Nguyen, B.-P. T. (2023). Ethical principles for
One Can Share? Uncovering Dimensions of
artificial intelligence in education. Education and
Collaborative Consumption of Consumer Goods.
Information Technologies, 28(4), 4221–4241.
International Textile and Apparel Association Annual
https://doi.org/10.1007/s10639-022-11316-w
Conference Proceedings, 75(1), Article 1.
https://www. [41] Nunes, J. C., Ordanini, A., & Giambastiani, G.
iastatedigitalpress.com/itaa/article/id/1451/ (2021). The Concept of Authenticity: What It Means
to Consumers. Journal of Marketing, 85(4), 1–20.
[31] Lefkeli, D., Karataş, M., & Gürhan-Canli, Z. (2024).
https://doi.org/10.1177/0022242921997
Sharing information with AI (versus a human)
081
impairs brand trust: The role of audience size
inferences and sense of exploitation. International [42] Rane, N. (2023). Enhancing customer loyalty through
Journal of Research in Marketing, 41(1), 138–155. Artificial Intelligence (AI), Internet of Things (IoT),
https://doi.org/ and Big Data technologies: Improving customer
10.1016/j.ijresmar.2023.08.011 satisfaction, engagement, relationship, and
experience. Internet of Things (IoT), and Big Data
[32] Lim, J. S., & Zhang, J. (2022). Adoption of AI-driven
Technologies: Improving Customer Satisfaction,
personalization in digital news platforms: An
Engagement, Relationship, and Experience (October
integrative model of technology acceptance and
13, 2023).
perceived contingency. Technology in Society, 69,
101965. [43] Reck, R., Castagna, A. C., Shuqair, S., & Costa Pinto,
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1545
D. (2022). The transparency paradox: When 542. https://doi.org/10.4300/JGME-5-4-18
transparency cues helps or backfires for brands?
[53] Sun, X., Pelet, J.-É., Dai, S., & Ma, Y. (2023). The
Journal of Cleaner Production, 372, 133381.
Effects of Trust, Perceived Risk, Innovativeness, and
https://doi.org/10.1016/j.jclepro.2022.133381
Deal Proneness on Consumers’ Purchasing Behavior
[44] Richey, R. G., Chowdhury, S., Davis‐Sramek, B., in the Livestreaming Social Commerce Context.
Giannakis, M., & Dwivedi, Y. K. (2023). Artificial Sustainability, 15(23), 16320. https://doi.
intelligence in logistics and supply chain org/10.3390/su152316320
management: A primer and roadmap for research.
[54] Tsai, C.-C., Lin, C.-L., & Chen, Y.-H. (2024). Impact
Journal of Business Logistics, 44(4), 532–549.
of Authenticity Perception on Experiential Value and
https://doi.org/10.
Customer Satisfaction under Contactless Services.
1111/jbl.12364
Systems, 12(1), 19.
[45] Sääksjärvi, M., & Lampinen, M. (2005). Consumer https://doi.org/10.3390/systems12010019
perceived risk in successive product generations.
[55] Verhoef, P. C., Stephen, A. T., Kannan, P., Luo, X.,
European Journal of Innovation Management, 8(2),
Abhishek, V., Andrews, M., Bart, Y., Datta, H.,
145–156. https://doi.org/10.1108/14601060
Fong, N., & Hoffman, D. L. (2017). Consumer
510594675
connectivity in a complex, technology-enabled, and
[46] Safeer, A. A., He, Y., Lin, Y., Abrar, M., & Nawaz, mobile-oriented world with smart products. Journal
Z. (2023). Impact of perceived brand authenticity on of Interactive Marketing, 40(1), 1–8.
consumer behavior: An evidence from generation Y
[56] Wamba-Taguimdje, S.-L., Fosso Wamba, S., Kala
in Asian perspective. International Journal of
Kamdjoug, J. R., & Tchatchouang Wanko, C. E.
Emerging Markets, 18(3), 685–704.
(2020). Influence of artificial intelligence (AI) on
https://doi.org/10.
firm performance: The business value of AI-based
1108/IJOEM-09-2020-1128
transformation projects. Business Process
[47] Sangtani, V., & Murshed, F. (2017). Product Management Journal, 26(7), 1893–1924.
knowledge and salesperson performance: Rethinking
[57] Wanner, J., Herm, L.-V., Heinrich, K., & Janiesch, C.
the role of optimism. Marketing Intelligence &
(2022). The effect of transparency and trust on
Planning, 35(6), 724–739. https://doi.
intelligent system acceptance: Evidence from a user-
org/10.1108/MIP-11-2016-0199
based study. Electronic Markets, 32(4), 2079–2102.
[48] Sohn, K., & Kwon, O. (2020). Technology https://doi.org/10.1007/s12525-022-00593-5
acceptance theories and factors influencing artificial
[58] Wu, L., Dodoo, N. A., & Wen, T. J. (2024).
Intelligence-based intelligent products. Telematics
Disclosing AI’s Involvement in Advertising to
and Informatics, 47, 101324.
Consumers: A Task-Dependent Perspective. Journal
[49] Spais, G., Phau, I., & Jain, V. (2023). AI marketing of Advertising, 1–19. https://doi.org/10.1080/
and AI ‐based promotions impact on consumer 00913367.2024.2309929
behavior and the avoidance of consumer autonomy
[59] Yang, H., Yu, J., Zo, H., & Choi, M. (2016). User
threat. Journal of Consumer Behaviour, cb.2248.
acceptance of wearable devices: An extended
https://doi.org/10.1002/cb.2248
perspective of perceived value. Telematics and
[50] Stampfl, R. W. (1978). Perceived Risk and Consumer Informatics, 33(2), 256–269.
Decision Making. Journal of Consumer Studies &
[60] Zerilli, J., Bhatt, U., & Weller, A. (2022). How
Home Economics, 2(3), 231–245.
transparency modulates trust in artificial intelligence.
https://doi.org/10.1111/j.1470-6431.1978.tb00324.x
Patterns, 3(4), 100455.
[51] Straub, E. T. (2009). Understanding technology https://doi.org/10.1016/j.patter.2022.100455
adoption: Theory and future directions for informal
learning. Review of Educational Research, 79(2),
625–649.
[52] Sullivan, G. M., & Artino, A. R. (2013). Analyzing
and Interpreting Data From Likert-Type Scales.
Journal of Graduate Medical Education, 5(4), 541–
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1546
Appendix
STUDY 1: Stimuli
Book Book Canvas Canvas
No AI Ai hint No AI Ai hint
STUDY 2: Stimuli
Human Human AI AI
Low information High information Low information High information
You are considering You are considering purchasing You are considering You are considering purchasing a
purchasing a new a new eBook titled 'Shadows of purchasing a new new eBook titled 'The Last
eBook titled 'Shadows Tomorrow', a gripping science eBook titled 'The Algorithm', an innovative thriller
of Tomorrow', a fiction novel acclaimed for its Last Algorithm', an that explores artificial
gripping science fiction deep storytelling and complex innovative thriller intelligence and the future of
novel. This novel has characters. This novel has been focusing on artificial humanity. This novel has been
been written by Alex R. written by Alex R. Johnson, a intelligence. This generated by an advanced AI
Johnson, a renowned renowned author known for novel has been program, designed to create
author known for his award-winning science fiction generated by an compelling narrative experiences.
contributions to the and fantasy novels. Before advanced AI Before making a decision, you
science fiction and making a decision, you have program. You have have access to a detailed
fantasy genres. You access to a detailed summary, access only to a brief summary, the table of contents,
have access only to a the table of contents, sample description of the sample chapters from various
brief description of the chapters from the beginning, book's premise, with parts of the book, and several
book's plot, without any middle, and end of the book, and no further positive reviews highlighting the
additional information, several positive reviews from information, sample novel's creativity and
sample chapters, or well-known book critics and chapters, or reviews engagement from technology
reviews readers alike. provided. enthusiasts and literary critics
STUDY 3: Stimuli
Artist Artist AI AI
Low outcome High outcome Low outcome High outcome
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(22s), 1530–1547 | 1547