Exploring fashion designers’ acceptance of AIGC: A dual-pathway analysis from the stimulus–organism–response perspective

Tingting Ma; Mengyun Yang

doi:10.1371/journal.pone.0356536

Abstract

Artificial Intelligence Generated Content (AIGC) is increasingly used in creative design. Understanding fashion designers’ willingness to adopt these tools has therefore become important for both research and practice. Drawing on the Stimulus–Organism–Response (SOR) model, this study integrates Self-Determination Theory (SDT) with perceived risk, social influence, and facilitating conditions. It examines how these contextual stimuli shape designers’ basic psychological need satisfaction and behavioral intention. We analyzed 318 valid responses with complete data for all 21 measurement items from Chinese fashion-design practitioners using partial least squares structural equation modeling (PLS-SEM). Perceived risk negatively predicted autonomy, competence, and relatedness, whereas social influence and facilitating conditions positively predicted these organismic states. Autonomy, competence, and relatedness each positively predicted behavioral intention, with competence showing the largest coefficient (β = 0.520, p < 0.001). Bootstrapped analyses confirmed all nine specific indirect effects from the three stimuli to behavioral intention through the three psychological needs. These findings clarify how SOR and SDT jointly explain technology adoption among fashion designers. They also provide practical guidance for copyright governance, prompt training, and collaboration between designers and AI.

Citation: Ma T, Yang M (2026) Exploring fashion designers’ acceptance of AIGC: A dual-pathway analysis from the stimulus–organism–response perspective. PLoS One 21(8): e0356536. https://doi.org/10.1371/journal.pone.0356536

Editor: Le Zhang, Sichuan University, CHINA

Received: January 16, 2026; Accepted: August 4, 2026; Published: August 17, 2026

Copyright: © 2026 Ma, Yang. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the manuscript and its Supporting Information files.

Funding: The author(s) received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

1 Introduction

With the rapid development of Artificial Intelligence Generated Content (AIGC) technology, the creative design industry is undergoing a transformation. Especially in the field of fashion design, AIGC technologies, such as Stable Diffusion and Midjourney, have been gradually embedded into pattern generation, style modeling, inspiration and other aspects, greatly enhancing design efficiency and freedom of expression. These technologies not only provide designers with more diversified creative tools, but also inject new momentum into the digital transformation of the apparel industry chain. Over the next three to five years, generative AI could add up to $150 billion at a conservative estimate, adding up to $275 billion to the operating profits of the apparel, fashion, and luxury industries [1]. However, despite its significant technological potential, AIGC adoption among fashion designers remains limited in practice. The low acceptance of AIGC by some designers [2] hides more complex psychological motivations and external environmental factors, which are worth exploring in depth.

Research on AIGC adoption has frequently drawn on the Technology Acceptance Model (TAM) [3], which emphasizes the influence of perceived usefulness and perceived ease of use. These evaluations are important, but a purely functional account is insufficient for fashion design, where adoption also depends on whether designers retain volition, feel capable of controlling the tool, and experience professional connection and support. SDT provides a basis for examining these organismic experiences through autonomy, competence, and relatedness [4,5]. These constructs represent basic psychological needs or need-satisfaction states; they are not direct measures of intrinsic motivation. SDT also does not by itself specify how negative and enabling features of the external environment enter the adoption process. Perceived Risk concerning originality, copyright, output controllability, and professional replacement therefore needs to be examined together with social and resource-based support.

To address this gap, the present study uses the SOR framework to organize a model in which perceived risk (PR), social influence (SI), and facilitating conditions (FC) act as external stimuli; autonomy (AUT), competence (COMP), and relatedness (REL) represent organismic basic-need-satisfaction states; and behavioral intention (BI) represents the response. PR captures uncertainty and possible losses associated with AIGC use. SI captures peer norms and organizational expectations, as conceptualized in the UTAUT framework [6]. FC captures the technical infrastructure, training, and assistance that reduce access barriers [7]. This mapping gives each construct a distinct theoretical role: SOR supplies the stimulus-to-organism-to-response structure, SDT specifies the organismic states, and technology-acceptance and perceived-risk research motivates the external predictors.

Using survey data from Chinese fashion-design practitioners, this study tests the direct paths from the three stimuli to autonomy, competence, and relatedness; the paths from these organismic states to behavioral intention; and the corresponding specific indirect effects. The study contributes by distinguishing basic psychological need satisfaction from intrinsic motivation, empirically testing the complete stimulus–organism–response transmission process, and identifying competence as the principal organismic predictor of AIGC acceptance. The findings also inform concrete practices for copyright governance, prompt-based training, and human-centered designer–AI collaboration.

2 Literature review and conceptual development

2.1 The rise and challenges of AIGC in fashion design

AIGC is one of the important breakthroughs in AI technology in recent years. By integrating basic algorithms such as Generative Adversarial Networks, Diffusion Models, and Transformer Architecture, the AIGC system is able to realize cross-modal and efficient content generation, including automatic creation of images, texts, videos, and 3D models. In this context, AIGC is gradually moving from a general-purpose content generation tool to deeply embedded applications in various industries, especially in creative industries such as visual design, media, advertising and fashion design [8].

In the field of fashion design, AIGC is widely used in many aspects of garment pattern generation, style simulation and effect creation. Designers can quickly generate conceptual drawings through text input, enhance design efficiency with AI, and optimize the ability to capture inspiration and visual expression [9]. In a study by Zhang & Liu [10], the findings suggest that AI tools can help fashion designers create visually expressive garments and ready-to-wear that meet the designers’ design criteria and consumers’ needs. AIGC shortens the time from concept to presentation and reduces the cost of product development compared to traditional rendering methods.

However, AIGC applications in fashion design also create practical concerns. Designers may question originality, copyright ownership, output controllability, stylistic convergence, and the possible devaluation of professional roles [11,12]. These concerns can weaken trust in AIGC and shape whether designers experience autonomy, competence, and relatedness when incorporating the technology into creative work. Accordingly, adoption depends not only on perceived functional advantages but also on basic psychological need satisfaction and perceptions of technological uncertainty. A multidimensional account of external context and organismic experience is therefore needed to explain designers’ acceptance decisions and inform the responsible integration of AIGC into fashion-design practice.

2.2 Theory of stimulus-organism-response (SOR)

The SOR model was originally proposed by environmental psychologists Mehrabian and Russell to explain how external environmental stimuli can trigger changes in an individual's internal mental state, which in turn leads to behavioral responses [13]. The model suggests that stimuli (S) are external environmental factors that trigger changes in the internal cognitive or affective state of an organism (O), resulting in a corresponding response (R). In recent years, the SOR model has been widely used in the fields of consumer behavior, e-commerce, and acceptance of virtual reality and artificial intelligence technologies [14–16]. With its clear structure, it can systematically show how the external environment influences individual behavior through psychological mechanisms, and is an important theoretical tool for studying the mechanism of user technology adoption.

The SOR framework has been explored by some scholars in AIGC research. For example, Gu et al. [17] explored consumers’ acceptance behavior of AI-generated advertisements based on the SOR model, and found that external advertisement features (e.g., presentation style, content quality) affect users’ cognitive and emotional states, which in turn influence their acceptance willingness. Zhou and Lu [18] introduced the SOR model in a digital library context and analyzed the mechanism of users’ trust in AIGC content. Although the application of SOR in the creative industry is relatively rare, Wang and Chen [19] attempted to integrate it with TAM and TPB to study graphic, product, and interaction design practitioners’ attitudes and willingness to adopt the AIGC tool, suggesting that the SOR structure is also applicable to modeling the psychology and behavior of creative workers. However, the above studies did not cover the fashion designer population, an industry with fewer studies on AIGC adoption.

In terms of motivational factors, some studies have incorporated psychological variables related to motivation, although they do not directly invoke the theory of motivation. For example, Wang and Chen [19] incorporated self-efficacy and subjective norms into the SOR model, which represent individual confidence and social pressure, respectively, and these factors can influence users’ motivation to adopt. Meanwhile, Liu et al. [20] introduced the emotional motivation variable “awe” based on the SOR model and found that positive awe (inspiration) enhanced users’ persuasive perceptions of AIGC content, while negative awe (threat) weakened this perception. These studies show that the SOR model is highly flexible and can be used in conjunction with perceived risk (e.g., trust, privacy and security) and motivation-related variables (e.g., emotion, self-efficacy) to expand its explanatory power. Although there are still few studies that directly integrate a complete motivation theory (e.g., SDT) into the SOR model, existing empirical evidence has validated that the framework can effectively cover the process of users’ psychological responses to AIGC technology.

2.3 Stimulus factors

2.3.1 Perceived risk (PR).

PR is defined as an individual's subjective perception of uncertainty about the possibility of suffering losses or adverse consequences when using a product or technology [21]. It is widely regarded as a key variable inhibiting users’ behavioral intention in technology adoption studies, especially in digital finance, online healthcare and AI scenarios, and its negative impact has been confirmed by a large number of empirical studies [22–24].

From the perspective of the SOR framework, Perceived risk can be conceptualized as an external stimulus that influences individuals’ internal cognitive or affective states and ultimately shapes their behavioral responses. Empirical evidence supports this positioning. For example, in the context of mobile health services, perceived risk has been shown to reduce continuance intention by weakening users’ trust in the platform [25]. Accordingly, treating perceived risk as an external stimulus is appropriate for explaining AIGC adoption in creative industries because it captures psychological responses to possible loss and uncertainty and clarifies how those responses may influence adoption decisions.

Perceived risk may affect behavioral intention indirectly by frustrating basic psychological need satisfaction. When risk is high, users may experience less control over technology use, which can reduce autonomy [26]. Uncertainty about operating AIGC can also reduce confidence in completing design tasks, weakening perceived competence. Concerns about originality, professional replacement, or unsafe outputs may further discourage interaction with other designers through the technology, weakening relatedness. Thus, perceived risk is expected to reduce all three organismic need-satisfaction states. This reasoning follows SDT, which identifies basic-need satisfaction as a condition for self-regulation [27].

Based on the above theoretical reasoning and empirical support, the following hypotheses are proposed:

H1a: Perceived risk negatively influences fashion designers’ perceived autonomy.
H1b: Perceived risk negatively influences fashion designers’ perceived competence.
H1c: Perceived risk negatively influences fashion designers’ perceived relatedness.

2.3.2 Social influence (SI).

SI refers to an individual's decision making that is influenced by the perceptions of others or the norms of a social group [6]. In the field of technology acceptance, social influence has been widely validated as one of the key variables in predicting behavioral intentions, especially when users lack familiarity or information asymmetry with the emerging technology, the guiding role of this external environmental factor is particularly significant [28]. Its mechanism of action includes both normative influences (individuals’ behavior due to catering to group expectations) and informational influences (individuals’ acquisition of knowledge and modification of attitudes through the experiences of others) [29].

In the context of creative designers’ AIGC use, social influence can support basic psychological need satisfaction through several routes. Recognition from colleagues, organizations, or professional communities can strengthen designers’ sense that AIGC use is self-endorsed, supporting autonomy. Observing others’ workflows and receiving prompt or tool guidance can lower the learning threshold and strengthen competence. Participation in AIGC-oriented groups and collaborative projects can create professional belonging and strengthen relatedness [4].

SDT proposes that environments supporting autonomy, competence, and relatedness facilitate more self-determined regulation and sustained engagement [27]. In this study, AUT, COMP, and REL are measured as separate organismic need-satisfaction constructs; intrinsic motivation is not measured directly. We therefore interpret the proposed paths as effects of need satisfaction on behavioral intention, rather than as a direct test of intrinsic motivation. Consistent with the SOR framework, no direct path from SI to BI is hypothesized because the model specifies transmission through organismic states.

H2a: Social influences positively affect fashion designers’ sense of autonomy.

H2b: Social influences positively affect fashion designers’ sense of competence.

H2c: Social influences positively affect fashion designers’ sense of relatedness.

2.3.3 Facilitating conditions (FC).

In technology adoption research, FC are defined as the availability of appropriate organizational and technical support when individuals perceive that they are using a system [6]. In AIGC-assisted fashion design, FC include suitable hardware and software, operational training, coworker assistance, and system stability. These resources can support autonomy by giving designers meaningful control over tool use, competence by helping them master prompt-based and iterative workflows, and relatedness by enabling collaboration and experience sharing [30–32].

FC therefore represents both an external resource and a context that can support basic psychological need satisfaction. When designers can obtain technical help, training, and peer support, they have more opportunities to explore AIGC voluntarily, master its functions, and participate in collaborative design work. Based on this reasoning, the following hypotheses are proposed:

H3a: Facilitating conditions positively influence fashion designers’ perceived autonomy in AIGC use.

H3b: Facilitating conditions positively influence fashion designers’ perceived competence in AIGC use.

H3c: Facilitating conditions positively influence fashion designers’ perceived relatedness in AIGC use.

2.4 Organismic states: Basic psychological needs

SDT identifies autonomy, competence, and relatedness as three basic psychological needs [33]. AUT refers to the feeling that one’s design choices are voluntary and self-endorsed. COMP refers to confidence in mastering the skills required to use AIGC and complete design tasks. REL refers to connection, recognition, and belonging within a community of designers. These constructs are distinct from intrinsic motivation: need satisfaction can support self-determined motivation, but the present questionnaire measures the three needs rather than motivation itself.

SDT predicts that social and technological environments supporting these needs promote internalized regulation and sustained engagement [34]. In creative work, autonomy supports authorship, competence supports confidence in iterative tool use, and relatedness supports exchange with peers and collaborators. The present model therefore places AUT, COMP, and REL in the organism layer and tests their separate effects on behavioral intention.

In the organism phase, this study examines whether designers feel autonomous when using AIGC, feel capable of operating the tools and completing design tasks, and experience connection and belonging with other designers through AIGC-related work. Prior evidence from AI-assisted design and other technology contexts shows that autonomy, competence, and relatedness can predict behavioral intention [35]. These findings support the use of SDT to specify the organismic mechanisms of AIGC adoption. Therefore, the following hypotheses are proposed in this study:

H4: Fashion designers’ perceived autonomy positively influences their behavioral intention toward AIGC use.

H5: Fashion designers’ perceived competence positively influences their behavioral intention toward AIGC use.

H6: Fashion designers’ perceived relatedness positively influences their behavioral intention toward AIGC use.

2.5 Conceptual model construction

Based on the above hypothesized paths, this study constructs the following conceptual model shown in Fig 1. SOR supplies the overall structure; PR, SI, and FC form the stimulus layer; AUT, COMP, and REL form the organismic basic-need-satisfaction layer; and BI forms the response layer. The model therefore contains seven core constructs and twelve hypothesized paths.

Download:

Fig 1. Conceptual model.

https://doi.org/10.1371/journal.pone.0356536.g001

3 Methods

3.1 Research design

This study used a quantitative, cross-sectional questionnaire design to test the theoretical model and hypotheses. The design measured perceived risk, social influence, facilitating conditions, three basic psychological need-satisfaction states, and behavioral intention in the natural work context of fashion-design practitioners. The study followed a deductive empirical approach: theory and prior evidence generated the hypotheses, and PLS-SEM tested the proposed relationships among the latent constructs [36].

3.2 Variable measurement and questionnaire design

To establish content validity, the questionnaire followed the scale-development process proposed by Churchill Jr. [37]. Perceived risk was adapted from Li et al. [11]; social influence and facilitating conditions were adapted from Venkatesh et al. [7]. The measurement items for AUT, COMP, and REL were adapted from Jiang et al. (2024) [4]. The research team adapted wording to the AIGC-assisted fashion-design context. The questionnaire contained 21 measurement items, with three reflective items for each construct, and used a seven-point Likert scale from 1 = strongly disagree to 7 = strongly agree. The full item wording and sources are provided in S1 Appendix.

A small pilot test was conducted before finalizing the questionnaire to further improve the validity of the instrument. The pre-survey invited approximately 20 volunteers with a background in the apparel design profession to complete the initial version of the questionnaire; these candidates had similar characteristics to the formal sample but were not included in the data analysis for the main study. The pre-survey collected feedback from the respondents on the difficulty of understanding the questions and the appropriateness of the wording, and the research team modified and improved the questionnaire accordingly. The results of the pre-survey showed that most of the questions were clear and easy to understand, and a few ambiguous expressions were optimized. Therefore, the formal questionnaire was finalized to contain 21 measurement items covering all the aforementioned constructs, with each construct being measured by three questions. Each item was measured on a seven-point Likert scale, i.e., from “1=Strongly Disagree” to “7=Strongly Agree”, and respondents rated each statement according to their own circumstances.

3.3 Common method bias

Because all variables were collected through the same self-report questionnaire, common-method bias was possible. The questionnaire clarified anonymity and academic use, separated predictor and outcome items, and avoided presenting all items in a single block. Harman’s one-factor test showed that the first unrotated factor explained 38.2% of the variance, below the 50% heuristic threshold [38]. This result reduces concern about a dominant common factor but does not eliminate all possible common-method bias.

3.4 Sample and participants

The respondents were apparel-design practitioners in China, covering groups such as apparel designers working in apparel companies, independent designers, college teachers of related majors, and senior students of apparel design majors. To ensure the professional relevance of the sample, all respondents were required to meet the following criteria: (1) experience in the field of apparel design or practice in the past 6 months; and (2) knowledge of or experience in the use of design tools based on AIGC. This standardized screening helped to improve the validity of the data, as only respondents with real exposure to apparel design and AIGC could provide meaningful answers to the relevant questions.

Before data collection, an a priori G*Power analysis used Cohen’s medium effect size (f² = 0.15), α = 0.05, power = 0.80, and three predictors, corresponding to the maximum number of structural paths directed at any endogenous construct; the minimum required sample size was 77 [39,40]. The final analytical dataset contained 318 valid response records with complete data across the 21 model items; therefore, all 318 records were used in the PLS-SEM analyses.

3.5 Data collection procedure

The data collection was mainly carried out during the period of February to March 2025. In order to improve the efficiency of questionnaire distribution and retrieval, this study utilized an online questionnaire, which was released through the Questionnaire Star platform. Online questionnaires have the advantages of low cost, wide coverage, and quick response to effectively reach geographically dispersed target groups [41]. The research team began by explaining the background and purpose of the study to potential respondents on the opening page of the questionnaire, emphasizing that participation was completely voluntary and responses were anonymous, and included screening questions to confirm that respondents met the requirements for experience in the apparel design industry and knowledge of the AIGC.

Participants were first presented with a detailed informed consent statement on the opening page of the questionnaire. They were required to indicate their voluntary participation by selecting a consent checkbox before proceeding. This procedure constituted electronic written informed consent. Only respondents who provided consent were allowed to continue with the survey, while those who declined were automatically exited. All participants were adults (aged 18 years and above), and no minors were involved in this study. This research was conducted in accordance with the principles of the Declaration of Helsinki. Formal ethical approval was not required because the study involved an anonymous online questionnaire, did not collect any personally identifiable or sensitive information, posed no foreseeable risk or harm to participants, and did not involve any intervention or deception. According to Article 32 of the Ethical Review Measures for Life Sciences and Medical Research Involving Humans issued by the National Health Commission of China (2023), studies that meet these conditions may be exempted from ethical review.

Only respondents who met the screening criteria completed the formal questionnaire. The survey link was distributed through design departments in apparel companies, universities with apparel-design programs, social-media channels, and professional designer communities. The final dataset contains 318 complete model questionnaires. Demographic information was available for 312 respondents; six respondents did not provide demographic items but had complete responses to all model items and were retained in the PLS-SEM analysis. Table 1 reports the demographic distributions using the 312 respondents with available demographic information.

Download:

Table 1. Demographic profile of respondents with available demographic information (n = 312).

https://doi.org/10.1371/journal.pone.0356536.t001

4 Results

SmartPLS version 4.1.0.3 was used to evaluate the measurement model and estimate the structural model using partial least squares structural equation modeling (PLS-SEM).

4.1 Measurement model

Before assessing the structural model, the reliability and convergent validity of the measurement model were evaluated (Table 2). Cronbach’s alpha values ranged from 0.828 to 0.933, rho_A values ranged from 0.881 to 0.934, and composite reliability values (rho_C) ranged from 0.893 to 0.957. Standardized indicator loadings ranged from 0.786 to 0.944, with all values exceeding the recommended threshold of 0.70. Average variance extracted (AVE) values ranged from 0.737 to 0.881 and exceeded the recommended threshold of 0.50. These results support the internal consistency reliability and convergent validity of the measurement model [42].

Download:

Table 2. Results of internal and convergent reliabilities.

https://doi.org/10.1371/journal.pone.0356536.t002

Discriminant validity was further assessed with the Fornell–Larcker criterion (Table 3). The square root of each construct’s AVE exceeded its correlations with the other constructs, supporting discriminant validity and the use of the measurement model in the structural analysis [42].

Download:

Table 3. Discriminant Validity (Fornell-Larcker).

https://doi.org/10.1371/journal.pone.0356536.t003

4.2 Structural model and hypothesis testing

The structural model was estimated with the 318 complete response records. The R² values for AUT, COMP, and REL were 0.445, 0.531, and 0.487, respectively, while the R² for BI was 0.696. The model therefore explained 69.6% of the variance in behavioral intention. VIF values ranged from 1.001 to 1.758, below the conservative threshold of 3.3, indicating no serious multicollinearity [43].

We tested the significance of the structural paths with 5,000 bootstrap resamples. Table 4 shows that PR negatively predicted AUT (β = −0.433, p < 0.001), COMP (β = −0.537, p < 0.001), and REL (β = −0.263, p < 0.001), supporting H1a–H1c. SI positively predicted AUT (β = 0.298, p < 0.001), COMP (β = 0.255, p < 0.001), and REL (β = 0.597, p < 0.001), supporting H2a–H2c. FC positively predicted AUT (β = 0.395, p < 0.001), COMP (β = 0.407, p < 0.001), and REL (β = 0.196, p < 0.001), supporting H3a–H3c.

Download:

Table 4. Structural model.

https://doi.org/10.1371/journal.pone.0356536.t004

Among the stimulus-to-organism paths, PR had the strongest negative association with COMP (β = −0.537***), whereas SI had the strongest positive association with REL (β = 0.597***). AUT (β = 0.313***), COMP (β = 0.520***), and REL (β = 0.136***) each positively predicted BI, supporting H4–H6. The complete structural model, including the standardized path coefficients, indicator loadings, and R² values of the endogenous constructs, is presented in Fig 2. The specific indirect effects are reported in Table 5. Every 95% bootstrap confidence interval excluded zero, confirming transmission from each stimulus to BI through each basic psychological need.

Download:

Table 5. Specific indirect effects of external stimuli on behavioral intention through basic psychological needs.

https://doi.org/10.1371/journal.pone.0356536.t005

Download:

Fig 2. Structural model path coefficient diagram (R² in circles).

https://doi.org/10.1371/journal.pone.0356536.g002

To further evaluate explanatory power and predictive relevance, Stone–Geisser’s Q² values and f² effect sizes were examined for the endogenous constructs and key structural paths. The effects of PR on COMP (f² = 0.609), SI on REL (f² = 0.689), COMP on BI (f² = 0.505), and FC on COMP (f² = 0.352) met or slightly exceeded the threshold for a large effect. The effect of PR on AUT (f² = 0.335) was medium [44]. These results further support the model’s explanatory strength.

All Q² values were greater than zero, with values of 0.604 for BI, 0.460 for COMP, 0.378 for AUT, and 0.387 for REL, indicating predictive relevance [44]. The comparatively high Q² values for BI and COMP suggest particularly strong predictive relevance for these endogenous constructs. Collectively, the path coefficients, R² values, effect sizes, and Q² values indicate satisfactory explanatory and predictive performance.

5 Discussion and conclusion

The SOR model shows that perceived risk, social influence, and facilitating conditions shape Chinese fashion designers’ AIGC adoption intention through distinct organismic states. PR reduced autonomy, competence, and relatedness, whereas SI and FC supported these states. These findings extend prior work by showing how risk, social norms, and resources enter the psychological process of technology adoption in fashion design.

At the organismic level, this study tested the three SDT basic psychological needs as separate constructs. AUT, COMP, and REL each positively predicted BI, so designers who felt more self-directed, capable, and connected reported stronger adoption intention. Because intrinsic motivation was not measured directly, the results do not establish that PR reduces intrinsic motivation or that SI and FC increase it; they establish the more specific pathway from contextual stimuli to need satisfaction and then to BI.

Competence had the strongest association with BI (β = 0.520, p < 0.001). Fashion designers must translate prompts into usable concepts, evaluate multiple outputs, control style and garment details, and integrate selected outputs into professional workflows. These tasks require practical tool mastery and confidence in making design judgments, so competence may be the most immediate condition for treating AIGC as a controllable collaborator rather than an unpredictable generator. The result is consistent with SDT’s emphasis on competence as a basis for sustained engagement [34] and with evidence that designers’ self-determination experiences shape their responses to AI participation [4].

In summary, the model identifies two complementary routes to AIGC adoption: perceived risk frustrates basic psychological need satisfaction, while social influence and facilitating conditions support it. The significant specific indirect effects show that the SOR sequence is empirically supported in this sample. The model should therefore be interpreted as a need-satisfaction mediation model, not as a direct measurement model of intrinsic motivation.

5.1 Comparison with related studies

To address possible overlap, we explicitly compared the present study with two related publications. Yang and Jin [45] analyzed 267 valid responses obtained from 335 collected questionnaires from Chinese fashion designers recruited through online design communities, professional associations, and social-media groups. The published article did not report the exact fieldwork dates. That study examined how personalization fit, perceived content quality, industry pressure, and perceived technological risk affected adoption intention through self-efficacy, innovativeness, and task–technology fit. Yang and Jin (2026) [5] collected 352 valid responses from January to February 2025 through the Questionnaire Star platform, recruiting participants through apparel companies and colleges and universities; it tested an SDT–TAM–perceived-risk model in which autonomy, competence, and relatedness affected perceived ease of use and perceived usefulness, which then predicted behavioral intention. By contrast, the present survey was separately administered in February–March 2025 using a different questionnaire and an independently exported 318-record analytical dataset. It tests whether perceived risk, social influence, and facilitating conditions affect behavioral intention through autonomy, competence, and relatedness within an SOR process. No response record from either published dataset was pooled with or reused in the present analysis. Because the three surveys were anonymous, participant-level identity matching was not possible. We therefore cannot exclude the possibility that some individuals may have independently participated in more than one survey.

5.2 Theoretical implications

This study integrates SOR with SDT by assigning each theory a distinct explanatory role. SOR specifies how external stimuli become organismic states and behavioral responses. SDT specifies the organismic states as autonomy, competence, and relatedness. Perceived-risk research motivates the inhibitory stimulus, while UTAUT motivates the social and resource-related stimuli. This bounded integration avoids treating risk, social influence, facilitating conditions, and psychological needs as interchangeable predictors and provides a clearer rationale for each hypothesized pathway.
The results extend SOR-based technology-acceptance research by testing both a negative stimulus and enabling stimuli in the same fashion-design model. PR was associated with lower need satisfaction, whereas SI and FC were associated with higher need satisfaction. The nine significant specific indirect effects further show how these external conditions are transmitted through AUT, COMP, and REL to BI. This evidence supports the use of SOR as a process model while preserving the conceptual distinction between SDT needs and intrinsic motivation.
The study also provides a contextualized account of AIGC adoption among Chinese fashion-design practitioners. Unlike generic technology-use models, the present model captures authorship and control through AUT, tool mastery through COMP, and professional community through REL. The strongest organismic path was COMP→BI, highlighting the importance of designers’ practical ability to direct, evaluate, and refine AIGC outputs in real design work.

5.3 Practical implications

The findings of this study have important practical implications for the popularization and application of AIGC in the field of creative design.

Perceived risk may hinder fashion designers from adopting AIGC technology, so all stakeholders should work to alleviate fashion designers’ concerns about risks and uncertainties. AIGC service providers and industry organizations can enhance the transparency and credibility of the tools by explaining matters such as copyright ownership, security, and privacy, thereby alleviating designers’ fears of unknown risks. For fashion design specifically, this includes clarifying ownership of AIGC-generated sketches and patterns, disclosing the training-data sources behind style generation to reduce the risk of inadvertent plagiarism of other designers’ work, and providing provenance or watermarking features so designers can document their own creative contribution for portfolio and rights-management purposes. This recommendation is consistent with Wang and Chen [19]. Companies should also emphasize that AIGC serves as a creative aid rather than a replacement for designers. In addition, offering targeted training and support can improve designers’ confidence and reduce insecurity. Prior studies have shown that privacy safeguards and learning resources are effective in lowering skepticism and increasing adoption rates [2]. As designers become more proficient with AIGC, their perceived risk tends to decline, increasing their willingness to adopt.
This study emphasizes the importance of supporting designers’ autonomy, competence, and relatedness during AIGC adoption. Organizations should preserve designers’ control over creative briefs, generation parameters, output selection, revision, and final approval; provide scaffolded training and feedback to strengthen competence; and create peer-learning communities and collaborative review routines to support relatedness. These conditions allow designers to experience AIGC as a tool they can direct and master rather than as an imposed system that threatens authorship or professional identity.
Because AIGC use in fashion design is largely prompt-driven, design firms and fashion-education programs should build structured prompt-engineering training that covers fabric, silhouette, and trend-specific vocabulary so designers can translate creative intent into usable outputs efficiently, rather than relying on trial-and-error. Workflow design should also formalize the designer-AI division of labor: designers set the creative brief, curate and iterate on AIGC outputs, and perform the final aesthetic and technical review, so that AIGC functions as a controllable collaborator embedded in professional practice rather than an autonomous decision-maker. Design schools and companies can further support this collaboration by building shared prompt libraries and case repositories that let designers learn from one another's successful AIGC workflows.
External incentives can be appropriately used to increase adoption rates, such as through industry awards or promoting designers who have used AIGC technology to produce outstanding design results, thereby creating a positive demonstration effect. When external motivation is internalized by individuals, it can also have a positive impact on behavior. Therefore, managers should focus on combining organizational goals with the personal values of designers to help them transform external requirements for using AIGC into internal recognition.

5.4 Limitations and future research

This study has four main limitations. First, the model measures autonomy, competence, and relatedness as basic psychological need-satisfaction constructs but does not directly measure intrinsic or extrinsic motivation; future studies should include validated motivation scales to test those mechanisms explicitly. Second, the cross-sectional design measures behavioral intention at one time point and cannot establish how intention develops into sustained use. Longitudinal or behavioral-log studies are needed. Third, the Chinese sample and the missing demographic responses for six participants limit generalizability and subgroup analysis. Fourth, the anonymous recruitment design prevents participant-level matching across related surveys; future multi-study programs should use ethically appropriate, non-identifying study codes to document overlap without compromising privacy.

Future research should compare designers across countries, organizational settings, and experience levels, and should examine whether copyright governance, prompt training, and collaborative tool design moderate the paths identified here. Experimental or longitudinal designs could also test whether gains in competence precede increases in actual AIGC use.

Supporting information

S1 Appendix. Measurement items and sources.

https://doi.org/10.1371/journal.pone.0356536.s001

(DOCX)

S1 Data. De-identified item-level survey data used in the PLS-SEM analysis.

https://doi.org/10.1371/journal.pone.0356536.s002

(XLSX)

Acknowledgments

The authors would like to thank the Nanchang Key Laboratory of Artificial Intelligence Manufacturing for Light Industrial Products for providing technical support and research facilities. We also sincerely appreciate all the participants who voluntarily took part in this study.

References

1. Harreis H, Koullias T, Roberts R, Te K. Generative AI in fashion | McKinsey. 2023 [cited 17 Apr 2025]. Available: https://www.mckinsey.com/industries/retail/our-insights/generative-ai-unlocking-the-future-of-fashion
- View Article
- Google Scholar
2. Yao Y, Wang X, Sun K. Research on key factors influencing Chinese designers’ use of AIGC: An extension based on TAM and TRI. PLoS One. 2025;20(2):e0314306. pmid:39970143
- View Article
- PubMed/NCBI
- Google Scholar
3. Davis FD. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quart. 1989;13:319–40.
- View Article
- Google Scholar
4. Jiang Q, Zang Y. To acknowledge or conceal: An exploratory study on designers’ self-determination factors and attitudes toward artificial intelligence participation in their works. Kybernetes. 2024; ahead of print.
- View Article
- Google Scholar
5. Yang M, Jin J. The adoption behavior of fashion designers toward AIGC: The interaction of psychological motivation and perceived risk. Acta Psychol (Amst). 2026;263:106194. pmid:41520450
- View Article
- PubMed/NCBI
- Google Scholar
6. Venkatesh V, Morris MG, Davis GB, Davis FD. User Acceptance of Information Technology: Toward A Unified View1. MIS Quart. 2003;27(3):425–78.
- View Article
- Google Scholar
7. Venkatesh V, Thong YL, Xu X. Consumer acceptance and use of information technology: Extending the unified theory of acceptance and use of technology. MIS Quart. 2012;36:157.
- View Article
- Google Scholar
8. Cao Y, Li S, Liu Y, Yan Z, Dai Y, Yu PS. A Comprehensive Survey of AI-Generated Content (AIGC): A History of Generative AI from GAN to ChatGPT. arXiv. 2023.
- View Article
- Google Scholar
9. Chen Y, Ma J. An intelligent generative method of fashion design combining attribute knowledge and Stable Diffusion Model. Textile Research Journal. 2024;95(11–12):1231–54.
- View Article
- Google Scholar
10. Zhang Y, Liu C. Unlocking the Potential of Artificial Intelligence in Fashion Design and E-Commerce Applications: The Case of Midjourney. J Theor Appl Electron Commer Res. 2024;19:654–70.
- View Article
- Google Scholar
11. Li X, Shen L, Ren X. Explore the Fashion Industry’s Behavioral Intention to Use Artificial Intelligence Generated Content Tools Based on the UTAUT Model. Int J Hum–Comput Interact. 2024;41(16):10285–300.
- View Article
- Google Scholar
12. Savas-Hall S, Koku PS, Mangleburg T. Really New Services: Perceived Risk and Adoption Intentions. Serv Mark Q. 2022;43:485–503.
- View Article
- Google Scholar
13. Mehrabian A, Russell JA. An approach to environmental psychology. Cambridge (MA): MIT Press; 1974.
14. Khosasih MM, Lisana L. Intention to Adopt Online Food Delivery Using Augmented Reality Mobile Apps: A Perspective of SOR Framework. Int J Adv Sci Eng Inf Technol. 2023;13(2):618–24.
- View Article
- Google Scholar
15. Kim MJ, Lee C-K, Jung T. Exploring Consumer Behavior in Virtual Reality Tourism Using an Extended Stimulus-Organism-Response Model. J Travel Res. 2020;59(1):69–89.
- View Article
- Google Scholar
16. Wu Z, Aw EC-X, Chuah SH-W. Webrooming as the new retail experience: a smart-shopping perspective. IJRDM. 2023;51(6):736–54.
- View Article
- Google Scholar
17. Gu C, Jia S, Lai J, Chen R, Chang X. Exploring Consumer Acceptance of AI-Generated Advertisements: From the Perspectives of Perceived Eeriness and Perceived Intelligence. JTAER. 2024;19(3):2218–38.
- View Article
- Google Scholar
18. Zhou T, Lu H. The effect of trust on user adoption of AI-generated content. EL. 2024;43(1):61–76.
- View Article
- Google Scholar
19. Wang S-F, Chen C-C. Explore the driving factors of designers’ AIGC usage behavior based on SOR framework. Front Comput Sci. 2024;6:1417016.
- View Article
- Google Scholar
20. Liu X, Yang B, Cheng H, Sun Y. Investigating the impact of AIGC features on the users’ perception of the persuasiveness of AIGC: A perspective of two-dimensional awe emotions. In: ICEB 2023 Proceedings (Chiayi, Taiwan). 2023. Available from: https://aisel.aisnet.org/iceb2023/41
21. Wang Y, Keller LR, Simon J. Descriptive models of perceived risk. In: Cochran JJ, Cox LA, Keskinocak P, Kharoufeh JP, Smith JC, editors. Wiley Encyclopedia of Operations Research and Management Science. Wiley; 2011. 10.1002/9780470400531.eorms0250
22. Wu T-C, Ho C-T. Reconstructing Risk Dimensions in Telemedicine: Investigating Technology Adoption and Barriers During the COVID-19 Pandemic in Taiwan. J Med Internet Res. 2025;27:e53306. pmid:39899842
- View Article
- PubMed/NCBI
- Google Scholar
23. Mehrolia S, Alagarsamy S, Moorthy V, Jeevananda S. Will Users Continue Using Banking Chatbots? The Moderating Role of Perceived Risk. FIIB Bus Rev. 2023;23197145231169900.
- View Article
- Google Scholar
24. Elasaria R, Nurabiah. Preference digital wallet by generation z with the unified theory of acceptance and use of technology 2 (utaut 2) model approach and perceived risk. EJMI. 2024;32:3–13.
- View Article
- Google Scholar
25. Jiang Y, Lau AKW. Understanding Post-Adoption Behavioral Intentions of Mobile Health Service Users: An Empirical Study during COVID-19. Int J Environ Res Public Health. 2023;20(5):3907. pmid:36900918
- View Article
- PubMed/NCBI
- Google Scholar
26. Kang H, Yang T, Banu N, Ng SWT, Lee JK. Exploring Smart Health Wearable Adoption Among Singaporean Older Adults Based on Self-Determination Theory: Web-Based Survey Study. JMIR Aging. 2025;8:e69008. pmid:40106231
- View Article
- PubMed/NCBI
- Google Scholar
27. Deci EL, Ryan RM. The “What” and “Why” of Goal Pursuits: Human Needs and the Self-Determination of Behavior. Psychol Inq. 2000;11(4):227–68.
- View Article
- Google Scholar
28. Rana MM, Siddiqee MS, Sakib MN, Ahamed MR. Assessing AI adoption in developing country academia: A trust and privacy-augmented UTAUT framework. Heliyon. 2024;10(18):e37569. pmid:39315142
- View Article
- PubMed/NCBI
- Google Scholar
29. Cheng M, Li X, Xu J. Promoting Healthcare Workers’ Adoption Intention of Artificial-Intelligence-Assisted Diagnosis and Treatment: The Chain Mediation of Social Influence and Human-Computer Trust. Int J Environ Res Public Health. 2022;19(20):13311. pmid:36293889
- View Article
- PubMed/NCBI
- Google Scholar
30. Briante F, Barabasch A. Supporting apprentices’ autonomy in vocational training: Insights into the practices at the Swiss Postal Service. JOVACET. 2024;7(1):21.
- View Article
- Google Scholar
31. Elmadag AB, Okan M, Kurtuldu E. Improving self-regulated learning competencies of service employees: roles of regulatory appraisals and supportive resources. J Organ Change Manag. 2022;36: 290–304.
- View Article
- Google Scholar
32. Walker R, Chong S, Chong J. Facilitating peer-led group research through virtual collaboration spaces: An exploratory research study. Res Learn Technol. 2021;29.
- View Article
- Google Scholar
33. Deci EL, Ryan RM. Self-determination theory: A macrotheory of human motivation, development, and health. Can Psychol. 2008;49(3):182–5.
- View Article
- Google Scholar
34. Ryan RM, Deci EL. Intrinsic and extrinsic motivation from a self-determination theory perspective: Definitions, theory, practices, and future directions. Contemp Educ Psychol. 2020;61:101860.
- View Article
- Google Scholar
35. Xie Y, Zhou R, Chan AHS, Jin M, Qu M. Motivation to interaction media: The impact of automation trust and self-determination theory on intention to use the new interaction technology in autonomous vehicles. Front Psychol. 2023;14:1078438. pmid:36844336
- View Article
- PubMed/NCBI
- Google Scholar
36. Creswell JW, Creswell JD. Research design: Qualitative, quantitative, and mixed methods approaches. 5th ed. Los Angeles: SAGE; 2018.
37. Churchill GA Jr. A Paradigm for Developing Better Measures of Marketing Constructs. J Market Res. 1979;16(1):64–73.
- View Article
- Google Scholar
38. Podsakoff PM, MacKenzie SB, Lee J-Y, Podsakoff NP. Common method biases in behavioral research: A critical review of the literature and recommended remedies. J Appl Psychol. 2003;88(5):879–903.
- View Article
- Google Scholar
39. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. New York: Routledge; 2013.
40. Faul F, Erdfelder E, Buchner A, Lang A-G. Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods. 2009;41(4):1149–60. pmid:19897823
- View Article
- PubMed/NCBI
- Google Scholar
41. Evans JR, Mathur A. The value of online surveys. Int Res. 2005;15(2):195–219.
- View Article
- Google Scholar
42. Fornell C, Larcker DF. Evaluating Structural Equation Models with Unobservable Variables and Measurement Error. J Market Res. 1981;18(1):39.
- View Article
- Google Scholar
43. Diamantopoulos A, Siguaw JA. Formative Versus Reflective Indicators in Organizational Measure Development: A Comparison and Empirical Illustration. Br J Manag. 2006;17(4):263–82.
- View Article
- Google Scholar
44. Hair JF, Risher JJ, Sarstedt M, Ringle CM. When to use and how to report the results of PLS-SEM. EBR. 2019;31(1):2–24.
- View Article
- Google Scholar
45. Yang M, Jin J. Analyzing the adoption of AIGC tools in fashion design: An S-O-R framework integrating task-technology fit. PLoS One. 2025;20(10):e0335522. pmid:41144471
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Harreis H, Koullias T, Roberts R, Te K. Generative AI in fashion | McKinsey. 2023 [cited 17 Apr 2025]. Available: https://www.mckinsey.com/industries/retail/our-insights/generative-ai-unlocking-the-future-of-fashion
View Article
Google Scholar

[2] View Article

[3] Google Scholar

[ref2] 2. Yao Y, Wang X, Sun K. Research on key factors influencing Chinese designers’ use of AIGC: An extension based on TAM and TRI. PLoS One. 2025;20(2):e0314306. pmid:39970143
View Article
PubMed/NCBI
Google Scholar

[5] View Article

[6] PubMed/NCBI

[7] Google Scholar

[ref3] 3. Davis FD. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quart. 1989;13:319–40.
View Article
Google Scholar

[9] View Article

[10] Google Scholar

[ref4] 4. Jiang Q, Zang Y. To acknowledge or conceal: An exploratory study on designers’ self-determination factors and attitudes toward artificial intelligence participation in their works. Kybernetes. 2024; ahead of print.
View Article
Google Scholar

[12] View Article

[13] Google Scholar

[ref5] 5. Yang M, Jin J. The adoption behavior of fashion designers toward AIGC: The interaction of psychological motivation and perceived risk. Acta Psychol (Amst). 2026;263:106194. pmid:41520450
View Article
PubMed/NCBI
Google Scholar

[15] View Article

[16] PubMed/NCBI

[17] Google Scholar

[ref6] 6. Venkatesh V, Morris MG, Davis GB, Davis FD. User Acceptance of Information Technology: Toward A Unified View1. MIS Quart. 2003;27(3):425–78.
View Article
Google Scholar

[19] View Article

[20] Google Scholar

[ref7] 7. Venkatesh V, Thong YL, Xu X. Consumer acceptance and use of information technology: Extending the unified theory of acceptance and use of technology. MIS Quart. 2012;36:157.
View Article
Google Scholar

[22] View Article

[23] Google Scholar

[ref8] 8. Cao Y, Li S, Liu Y, Yan Z, Dai Y, Yu PS. A Comprehensive Survey of AI-Generated Content (AIGC): A History of Generative AI from GAN to ChatGPT. arXiv. 2023.
View Article
Google Scholar

[25] View Article

[26] Google Scholar

[ref9] 9. Chen Y, Ma J. An intelligent generative method of fashion design combining attribute knowledge and Stable Diffusion Model. Textile Research Journal. 2024;95(11–12):1231–54.
View Article
Google Scholar

[28] View Article

[29] Google Scholar

[ref10] 10. Zhang Y, Liu C. Unlocking the Potential of Artificial Intelligence in Fashion Design and E-Commerce Applications: The Case of Midjourney. J Theor Appl Electron Commer Res. 2024;19:654–70.
View Article
Google Scholar

[31] View Article

[32] Google Scholar

[ref11] 11. Li X, Shen L, Ren X. Explore the Fashion Industry’s Behavioral Intention to Use Artificial Intelligence Generated Content Tools Based on the UTAUT Model. Int J Hum–Comput Interact. 2024;41(16):10285–300.
View Article
Google Scholar

[34] View Article

[35] Google Scholar

[ref12] 12. Savas-Hall S, Koku PS, Mangleburg T. Really New Services: Perceived Risk and Adoption Intentions. Serv Mark Q. 2022;43:485–503.
View Article
Google Scholar

[37] View Article

[38] Google Scholar

[ref13] 13. Mehrabian A, Russell JA. An approach to environmental psychology. Cambridge (MA): MIT Press; 1974.

[ref14] 14. Khosasih MM, Lisana L. Intention to Adopt Online Food Delivery Using Augmented Reality Mobile Apps: A Perspective of SOR Framework. Int J Adv Sci Eng Inf Technol. 2023;13(2):618–24.
View Article
Google Scholar

[41] View Article

[42] Google Scholar

[ref15] 15. Kim MJ, Lee C-K, Jung T. Exploring Consumer Behavior in Virtual Reality Tourism Using an Extended Stimulus-Organism-Response Model. J Travel Res. 2020;59(1):69–89.
View Article
Google Scholar

[44] View Article

[45] Google Scholar

[ref16] 16. Wu Z, Aw EC-X, Chuah SH-W. Webrooming as the new retail experience: a smart-shopping perspective. IJRDM. 2023;51(6):736–54.
View Article
Google Scholar

[47] View Article

[48] Google Scholar

[ref17] 17. Gu C, Jia S, Lai J, Chen R, Chang X. Exploring Consumer Acceptance of AI-Generated Advertisements: From the Perspectives of Perceived Eeriness and Perceived Intelligence. JTAER. 2024;19(3):2218–38.
View Article
Google Scholar

[50] View Article

[51] Google Scholar

[ref18] 18. Zhou T, Lu H. The effect of trust on user adoption of AI-generated content. EL. 2024;43(1):61–76.
View Article
Google Scholar

[53] View Article

[54] Google Scholar

[ref19] 19. Wang S-F, Chen C-C. Explore the driving factors of designers’ AIGC usage behavior based on SOR framework. Front Comput Sci. 2024;6:1417016.
View Article
Google Scholar

[56] View Article

[57] Google Scholar

[ref20] 20. Liu X, Yang B, Cheng H, Sun Y. Investigating the impact of AIGC features on the users’ perception of the persuasiveness of AIGC: A perspective of two-dimensional awe emotions. In: ICEB 2023 Proceedings (Chiayi, Taiwan). 2023. Available from: https://aisel.aisnet.org/iceb2023/41

[ref21] 21. Wang Y, Keller LR, Simon J. Descriptive models of perceived risk. In: Cochran JJ, Cox LA, Keskinocak P, Kharoufeh JP, Smith JC, editors. Wiley Encyclopedia of Operations Research and Management Science. Wiley; 2011. 10.1002/9780470400531.eorms0250

[ref22] 22. Wu T-C, Ho C-T. Reconstructing Risk Dimensions in Telemedicine: Investigating Technology Adoption and Barriers During the COVID-19 Pandemic in Taiwan. J Med Internet Res. 2025;27:e53306. pmid:39899842
View Article
PubMed/NCBI
Google Scholar

[61] View Article

[62] PubMed/NCBI

[63] Google Scholar

[ref23] 23. Mehrolia S, Alagarsamy S, Moorthy V, Jeevananda S. Will Users Continue Using Banking Chatbots? The Moderating Role of Perceived Risk. FIIB Bus Rev. 2023;23197145231169900.
View Article
Google Scholar

[65] View Article

[66] Google Scholar

[ref24] 24. Elasaria R, Nurabiah. Preference digital wallet by generation z with the unified theory of acceptance and use of technology 2 (utaut 2) model approach and perceived risk. EJMI. 2024;32:3–13.
View Article
Google Scholar

[68] View Article

[69] Google Scholar

[ref25] 25. Jiang Y, Lau AKW. Understanding Post-Adoption Behavioral Intentions of Mobile Health Service Users: An Empirical Study during COVID-19. Int J Environ Res Public Health. 2023;20(5):3907. pmid:36900918
View Article
PubMed/NCBI
Google Scholar

[71] View Article

[72] PubMed/NCBI

[73] Google Scholar

[ref26] 26. Kang H, Yang T, Banu N, Ng SWT, Lee JK. Exploring Smart Health Wearable Adoption Among Singaporean Older Adults Based on Self-Determination Theory: Web-Based Survey Study. JMIR Aging. 2025;8:e69008. pmid:40106231
View Article
PubMed/NCBI
Google Scholar

[75] View Article

[76] PubMed/NCBI

[77] Google Scholar

[ref27] 27. Deci EL, Ryan RM. The “What” and “Why” of Goal Pursuits: Human Needs and the Self-Determination of Behavior. Psychol Inq. 2000;11(4):227–68.
View Article
Google Scholar

[79] View Article

[80] Google Scholar

[ref28] 28. Rana MM, Siddiqee MS, Sakib MN, Ahamed MR. Assessing AI adoption in developing country academia: A trust and privacy-augmented UTAUT framework. Heliyon. 2024;10(18):e37569. pmid:39315142
View Article
PubMed/NCBI
Google Scholar

[82] View Article

[83] PubMed/NCBI

[84] Google Scholar

[ref29] 29. Cheng M, Li X, Xu J. Promoting Healthcare Workers’ Adoption Intention of Artificial-Intelligence-Assisted Diagnosis and Treatment: The Chain Mediation of Social Influence and Human-Computer Trust. Int J Environ Res Public Health. 2022;19(20):13311. pmid:36293889
View Article
PubMed/NCBI
Google Scholar

[86] View Article

[87] PubMed/NCBI

[88] Google Scholar

[ref30] 30. Briante F, Barabasch A. Supporting apprentices’ autonomy in vocational training: Insights into the practices at the Swiss Postal Service. JOVACET. 2024;7(1):21.
View Article
Google Scholar

[90] View Article

[91] Google Scholar

[ref31] 31. Elmadag AB, Okan M, Kurtuldu E. Improving self-regulated learning competencies of service employees: roles of regulatory appraisals and supportive resources. J Organ Change Manag. 2022;36: 290–304.
View Article
Google Scholar

[93] View Article

[94] Google Scholar

[ref32] 32. Walker R, Chong S, Chong J. Facilitating peer-led group research through virtual collaboration spaces: An exploratory research study. Res Learn Technol. 2021;29.
View Article
Google Scholar

[96] View Article

[97] Google Scholar

[ref33] 33. Deci EL, Ryan RM. Self-determination theory: A macrotheory of human motivation, development, and health. Can Psychol. 2008;49(3):182–5.
View Article
Google Scholar

[99] View Article

[100] Google Scholar

[ref34] 34. Ryan RM, Deci EL. Intrinsic and extrinsic motivation from a self-determination theory perspective: Definitions, theory, practices, and future directions. Contemp Educ Psychol. 2020;61:101860.
View Article
Google Scholar

[102] View Article

[103] Google Scholar

[ref35] 35. Xie Y, Zhou R, Chan AHS, Jin M, Qu M. Motivation to interaction media: The impact of automation trust and self-determination theory on intention to use the new interaction technology in autonomous vehicles. Front Psychol. 2023;14:1078438. pmid:36844336
View Article
PubMed/NCBI
Google Scholar

[105] View Article

[106] PubMed/NCBI

[107] Google Scholar

[ref36] 36. Creswell JW, Creswell JD. Research design: Qualitative, quantitative, and mixed methods approaches. 5th ed. Los Angeles: SAGE; 2018.

[ref37] 37. Churchill GA Jr. A Paradigm for Developing Better Measures of Marketing Constructs. J Market Res. 1979;16(1):64–73.
View Article
Google Scholar

[110] View Article

[111] Google Scholar

[ref38] 38. Podsakoff PM, MacKenzie SB, Lee J-Y, Podsakoff NP. Common method biases in behavioral research: A critical review of the literature and recommended remedies. J Appl Psychol. 2003;88(5):879–903.
View Article
Google Scholar

[113] View Article

[114] Google Scholar

[ref39] 39. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. New York: Routledge; 2013.

[ref40] 40. Faul F, Erdfelder E, Buchner A, Lang A-G. Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods. 2009;41(4):1149–60. pmid:19897823
View Article
PubMed/NCBI
Google Scholar

[117] View Article

[118] PubMed/NCBI

[119] Google Scholar

[ref41] 41. Evans JR, Mathur A. The value of online surveys. Int Res. 2005;15(2):195–219.
View Article
Google Scholar

[121] View Article

[122] Google Scholar

[ref42] 42. Fornell C, Larcker DF. Evaluating Structural Equation Models with Unobservable Variables and Measurement Error. J Market Res. 1981;18(1):39.
View Article
Google Scholar

[124] View Article

[125] Google Scholar

[ref43] 43. Diamantopoulos A, Siguaw JA. Formative Versus Reflective Indicators in Organizational Measure Development: A Comparison and Empirical Illustration. Br J Manag. 2006;17(4):263–82.
View Article
Google Scholar

[127] View Article

[128] Google Scholar

[ref44] 44. Hair JF, Risher JJ, Sarstedt M, Ringle CM. When to use and how to report the results of PLS-SEM. EBR. 2019;31(1):2–24.
View Article
Google Scholar

[130] View Article

[131] Google Scholar

[ref45] 45. Yang M, Jin J. Analyzing the adoption of AIGC tools in fashion design: An S-O-R framework integrating task-technology fit. PLoS One. 2025;20(10):e0335522. pmid:41144471
View Article
PubMed/NCBI
Google Scholar

[133] View Article

[134] PubMed/NCBI

[135] Google Scholar

Figures

Abstract

1 Introduction

2 Literature review and conceptual development

2.1 The rise and challenges of AIGC in fashion design

2.2 Theory of stimulus-organism-response (SOR)

2.3 Stimulus factors

2.3.1 Perceived risk (PR).

2.3.2 Social influence (SI).

2.3.3 Facilitating conditions (FC).

2.4 Organismic states: Basic psychological needs

2.5 Conceptual model construction

3 Methods

3.1 Research design

3.2 Variable measurement and questionnaire design

3.3 Common method bias

3.4 Sample and participants

3.5 Data collection procedure

4 Results

4.1 Measurement model

4.2 Structural model and hypothesis testing

5 Discussion and conclusion

5.1 Comparison with related studies

5.2 Theoretical implications

5.3 Practical implications

5.4 Limitations and future research

Supporting information

S1 Appendix. Measurement items and sources.

S1 Data. De-identified item-level survey data used in the PLS-SEM analysis.

Acknowledgments

References