When AI Fairness Shapes Creativity: The Mediating Role of Attitudes Toward AI Across Gender

Abstract

Artificial intelligence (AI) is transforming the modern workplace by offering unprecedented opportunities to enhance employee creativity and organizational innovation. In the context of digital transformation, organizations are striving to ensure sustainable performance; however, research remains limited on how perceived AI fairness and attitudes toward AI jointly influence creativity. Grounded in Social Exchange Theory and the Technology Acceptance Model, this study proposes a moderated mediation model to examine how perceived AI fairness shapes employees’ attitudes toward AI and, in turn, their creativity, with gender acting as a moderator of the relationship between fairness perceptions and attitudes toward AI. Data were collected from 214 highly skilled employees from diverse cultural backgrounds working in technologically advanced environments. Using Partial Least Squares Structural Equation Modeling (PLS-SEM), the findings reveal a positive association between perceived AI fairness and creativity. Attitudes toward AI partially mediate this relationship; however, gender does not exert a significant moderating effect. The findings highlight the importance of AI fairness, reinforced by positive attitudes toward AI, in enhancing employee creativity. They also underscore the need for responsible and equitable AI practices and provide context-specific insights into the ethical challenges of AI in socio-technologically vulnerable environments. Finally, the findings point to a shift toward a more egalitarian and inclusive organizational landscape, in which gender differences become less salient in the context of digital transformation.

1. Introduction

In the current context of rapid digital transformation, organizations are increasingly adopting AI to foster innovation and strengthen their competitive advantage (Y. Zhou et al., 2026; DiClaudio, 2019). AI refers to computer systems capable of performing complex tasks traditionally associated with human intelligence, such as reasoning, decision-making, and problem-solving, by drawing on models trained on existing data (Boussioux et al., 2024). Although AI is often portrayed as a potential driver of creativity (Ivcevic & Grandinetti, 2024), its actual impact remains contested and inconclusive (Sun et al., 2025; Gallup Survey, 2025). This ambiguity highlights an important research gap and underscores the need to examine more closely the conditions under which AI enhances or constrains human creativity.

Recent studies have emphasized the importance of investigating organizational factors to better understand how individuals make sense of AI use (Yoo, 2025; Daly et al., 2025; Marocco et al., 2025; Lichtenthaler, 2020). One particularly promising organizational driver is perceived AI fairness (PAF), which has been identified as crucial for enhancing organizational performance while reducing uncertainty and alleviating anxiety (Colquitt et al., 2023; Yoo, 2025; Köchling et al., 2025; Amari, 2024; Amari et al., 2024). PAF refers to employees’ general perception that their organization applies AI-driven practices in a fair, transparent, and consistent manner. PAF is an important point of interest in this study since the widespread adoption of AI raises critical ethical concerns regarding the future of work practices (Hechler et al., 2025; Brem & Hörauf, 2025; Koenig, 2025; Tambe et al., 2019). AI adoption has been shown to exacerbate organizational inequalities and generate new forms of bias and discrimination (Bankins et al., 2024; Tambe et al., 2019). Such dynamics may increase perceptions of injustice, thereby fostering apprehension and resistance toward AI and limiting employees’ engagement and trust in these technologies (Yoo, 2025). It was found that when employees believe that their organization treats them fairly while using AI, they tend to exhibit greater enthusiasm and deeper involvement in their creative tasks (Y. Zhou et al., 2026; El Achi et al., 2026; Ma et al., 2026).

To gain deeper insight into how individuals’ interpretations of AI fairness translate into creativity, attitudes toward AI (AAI) are examined as a potential mediating mechanism. AAI can be understood as a psychological disposition through which employees evaluate these technologies with varying degrees of acceptance, enthusiasm, skepticism, or anxiety (Schepman & Rodway, 2020, 2023). This decision is motivated by the fact that the effectiveness of AI in organizational settings depends not only on its technical capabilities but also on how employees perceive and make sense of these technologies (Savolainen et al., 2026; Park et al., 2024).

In addition, individual differences, particularly gender, may shape both employees’ perceptions of PAF (Colquitt et al., 2023) and their use of AI. This perspective further underscores the importance of considering gender when examining technology-related perceptions and responses (Hechler et al., 2025; Schepman & Rodway, 2020, 2023). However, empirical findings remain inconclusive, and evidence regarding gender differences in the relationship between PAF and AAI remains fragmented. Against this backdrop, the present study introduces gender as a boundary condition to provide a more nuanced understanding of how differences in perceptions of AI fairness translate into favorable attitudes toward AI and, ultimately, enhanced creative performance. Accordingly, this study asks: How does PAF influence employees’ AAI, thereby fostering creative thinking, and to what extent is this relationship moderated by gender?

In addressing these questions, this study foregrounds the role of PAF in shaping technology-driven outcomes, particularly creativity which is described in this study as a distinctly human process that can be enhanced through collaboration with AI (Runco, 2023). First, this study responds to Colquitt et al. (2023) call to examine how PAF shapes individuals’ outcomes while facing emerging technologies. Thus, the relationship between PAF and creativity represents a promising area of inquiry, particularly in light of limited findings regarding the impact of PAF on creativity in the context of AI (Köchling et al., 2025; Yoo, 2025). In this vein, Social Exchange Theory (SET) is mobilized to suggest that when employees perceive AI-related practices as fair, transparent, and ethically grounded, they are more likely to develop positive AAI and experience heightened intrinsic motivation. In turn, in accordance with reciprocity norms, these positive psychological states are expected to translate into behaviors that benefit the organization through enhanced creative performance (Amari et al., 2024; Amabile, 1996). Second, by examining employees’ AAI as a key mediating mechanism between PAF and creativity, this study offers novel perspectives into the psychological processes through which fairness perceptions influence AI-driven creative outcomes (Savolainen et al., 2026; Ma et al., 2026). To do so, this study draws on the Technology Acceptance Model (TAM) to better capture AAI’s outcomes. Third, this study advances the literature on gender equality, by explicitly examining gender as a moderator. It provides novel insights into how gender norms shape employees’ PAF, particularly in relation to their creative engagement and subsequent performance outcomes. Lastly, most research on human–AI interaction and ethical perceptions has focused on specific, often monocultural, contexts. This study addresses this gap by examining the relationship between PAF, attitudes toward AI, and creativity in AI-enabled settings from a broader perspective. In doing so, it includes participants from diverse cultural backgrounds, thereby yielding more nuanced findings. Practically, the findings provide actionable insights for organizations seeking to design fair, ethical, and inclusive AI systems that foster employee engagement and creativity, particularly in highly challenging contexts.

2. Theoretical Framework

2.1. Fairness and Creativity in AI Contexts

The extensive debate on AI ethics has attracted attention, particularly regarding its effect on organizational behavior outcomes, especially on creativity (Yoo, 2025; Köchling et al., 2025; Tambe et al., 2019). Creativity is commonly defined as the ability to generate ideas or solutions that are both novel and useful, thereby fostering innovation within organizations (J. Zhou & George, 2001; Amabile, 1996).

In the context of digital transformation, the contours of creativity are evolving, paving the way for different interpretations in the literature. Runco (2023) argues that Generative AI systems such as ChatGPT do not and will never possess the capacity to be creative in the human sense. According to this perspective, what may be obtained through these systems corresponds rather to a form of “artificial creativity,” since these technologies operate as reactive systems that generate automated responses based on user prompts, without intrinsic purpose, intentionality, or autonomous creative agency (Runco, 2023). More recently, creativity has been conceptualized as the outcome of a collaborative interaction between humans and intelligent systems, in which each contributes complementary capabilities to the creative process (Brem & Hörauf, 2025; Grilli & Pedota, 2024). This interaction may support different manifestations of creativity, including idea generation, problem-solving, and creative risk-taking (Sun et al., 2025; Dong et al., 2025). However, such perspectives do not necessarily imply that AI itself is creative. Rather, creativity in this study is understood as a fundamentally human creative process that can be enhanced through human–AI collaboration. In line with Eapen et al. (2023), creativity corresponds to an “augmented creativity” process in which AI supports human intelligence in co-developing the best solution without substituting human thinking. More specifically, this form of augmented creativity emerges from a deliberate and reflective collaboration with AI, where the employee remains central to the creative process through the interpretation, evaluation, selection, modification, and transformation of AI-generated content into meaningful and useful outcomes (Eapen et al., 2023). Therefore, creativity is not equated with the mere production of multiple AI-generated ideas or increased idea productivity. Instead, the creative contribution resides in the employee’s judgment, contextual understanding, and purposeful application of AI-generated suggestions. As such, AI becomes a technological–cognitive support capable of stimulating divergent thinking and enhancing the individual’s creative potential.

Nevertheless, there are divergences regarding the influence of AI on creativity. While AI can act as a powerful catalyst for ideation and support creative processes (Ma et al., 2026; Boussioux et al., 2024; Doshi & Hauser, 2024), it may also hinder creativity by promoting homogeneous outputs, reducing independent thinking, and encouraging excessive reliance on technological tools (Khoso et al., 2026; Brem & Hörauf, 2025; Dong et al., 2025; Raisch & Krakowski, 2021). This finding motivates our contribution to this debate and our interest in examining how to make AI-assisted creativity more ethical and sustainable. In this vein, researchers advocate that PAF could be an enabler for AI-augmented creativity (Yoo, 2025; Colquitt et al., 2023).

It is worthwhile to recall in this respect that, in this study, PAF is conceptualized as employees’ general belief that their organization implements AI-driven practices in a fair and transparent manner, ensuring equitable treatment within the organization (Amari et al., 2024; Ambrose & Schminke, 2009). While some studies focus on specific aspects of AI fairness (Y. Zhou et al., 2026), this study adopts a holistic perspective, thereby providing a comprehensive foundation for understanding the impact of AI fairness on employees’ attitudes and creative behavior (Ambrose & Schminke, 2009). In practice, AI fairness could be manifested through transparency, absence of bias, and consistency likely to shape employees’ engagement in AI-augmented tasks (Yoo, 2025; Köchling et al., 2025).

Prior research indicates a positive association between PAF and creative behaviors (Khoso et al., 2026; Tran Pham, 2023). When AI systems are perceived as fair and ethically grounded, they tend to foster trust, engagement, and psychological safety among employees (Amari et al., 2026; Amari, 2023). This, in turn, encourages employees to collaborate more meaningfully with AI tools (El Achi et al., 2026; Yoo, 2025; Raisch & Krakowski, 2021). More importantly, fair AI practices support employees’ reskilling and upskilling by facilitating learning and adaptation to evolving work demands. As a result, employees develop greater confidence and stronger capabilities in using AI for problem-solving (World Economic Forum, 2024; Raisch & Krakowski, 2021). Ethical AI practices also reassure employees by signaling that AI is intended to support their work and enhance their capabilities rather than control, replace, or unfairly evaluate them (Raisch & Krakowski, 2021).

To further strengthen these arguments, SET offers a useful theoretical lens for understanding why perceived AI fairness fosters creative behaviors (Cropanzano & Mitchell, 2005). According to the principle of reciprocity, when organizations implement AI-based practices perceived as supportive and fair, employees’ intrinsic motivation increases (Colquitt et al., 2001). This, in turn, creates conditions in which employees feel psychologically safe. They are more willing to experiment, take risks, and generate novel and useful ideas (Amari et al., 2024; Amabile & Pratt, 2016; Edmondson, 1999). In other words, when AI practices are perceived as fair and transparent, concerns about algorithmic opacity or bias are reduced. Employees are then more likely to reciprocate this positive treatment. They invest more deeply in their work and show higher engagement in creative tasks (Y. Zhou et al., 2026; Yoo, 2025; Amari et al., 2024; Tambe et al., 2019; Zhang & Bartol, 2010). Conversely, when AI-driven practices are perceived as unfair, employees may feel neglected. In such cases, injustice acts as a “corrosive solvent” (Cropanzano et al., 2007). It undermines organizational relationships and discourages creative engagement (Yoo, 2025; Colquitt et al., 2013). Based on these insights and SET lens, PAF is expected to foster employees’ creative behavior. Accordingly, the following hypothesis is proposed:

H1. 

Perceived AI fairness positively influences employees’ creativity.

2.2. Perceived AI Fairness, Gender and Attitudes Toward AI

Nascent empirical evidence suggests that individuals’ interpretations of AI fairness serve as a key cognitive lens through which employees make sense of AI-driven practices (Yoo, 2025; Köchling et al., 2025). Indeed, employees actively construct meaning around AI systems based on their perceptions of fairness rather than passively accepting their outcomes (Yoo, 2025). Even when AI-based decisions are objectively or statistically fair, employees may still perceive them as unfair, reflecting the socially constructed nature of fairness perceptions (Cropanzano et al., 2007, 2017). Accordingly, perceived AI fairness shapes how employees cognitively frame and evaluate AI and influences their attitudinal responses.

AAI refers to a psychological tendency to evaluate these technologies positively or negatively. They can take the form of acceptance, enthusiasm, skepticism, or fear (Schepman & Rodway, 2020, 2023). They also foster positive attitudes and behaviors such as creativity (Yoo, 2025; Köchling et al., 2025). Prior research shows that when individuals perceive AI-driven organizational decisions as fair and transparent, they feel more confident. This reduces anxiety and apprehension and fosters greater acceptance of these technologies (Köchling et al., 2025; Yoo, 2025; Marocco et al., 2025; Tambe et al., 2019). Conversely, when algorithmic opacity and decision-making criteria are unclear or unjustified, employees are more likely to resist. They may feel frustrated and skeptical, particularly when automated decisions lack adequate explanation (Köchling et al., 2025; Yoo, 2025; Tambe et al., 2019).

The relationship between PAF and employees’ AAI can be further understood through the lens of SET. Drawing on a cost–benefit perspective, consistent with equity theory (Adams, 1963), individuals tend to evaluate the perceived benefits of AI against its potential costs. Benefits include increased efficiency, workload reduction, and performance improvement while costs include privacy concerns, algorithmic bias, job insecurity, and displacement. When perceived benefits outweigh the costs, employees are more likely to perceive AI systems as fair. This fosters more positive attitudes and stronger intentions to adopt AI (El Achi et al., 2026; Daly et al., 2025; Yoo, 2025; Marocco et al., 2025). Conversely, when perceived costs dominate, feelings of injustice may emerge, leading to resistance toward AI (Köchling et al., 2025; Yoo, 2025; Daly et al., 2025; Tambe et al., 2019). Accordingly, based on these arguments, PAF is expected to stimulate positive AAI among individuals. Thus, the following hypothesis is proposed:

H2. 

Perceived AI fairness positively influences employees’ attitudes toward AI.

Researchers have consistently argued that perceptions of fairness are not uniform but instead vary across individuals (Colquitt et al., 2023), with gender emerging as a salient differentiating factor. Prior research indicates that individuals from historically marginalized groups, particularly women, exhibit heightened sensitivity to potential biases embedded in AI systems (West et al., 2019). This heightened sensitivity is reflected in a more critical evaluation of the equity of algorithmic outcomes. Women are more likely to perceive AI systems as less fair and to scrutinize the processes through which decisions are generated. In this regard, procedural elements such as transparency assume particular importance. Transparency is especially critical in AI contexts, where decision-making processes are often opaque and susceptible to embedded biases (Schepman & Rodway, 2020, 2023). Women tend to place greater emphasis on the visibility and explainability of these processes, reflecting broader concerns about accountability and ethical integrity. This orientation aligns with evidence suggesting that women demonstrate stronger concern for the societal and ethical implications of AI technologies. By contrast, men are generally found to prioritize structural and procedural efficiency over ethical scrutiny (Schepman & Rodway, 2020, 2023).

Additionally, existing studies further suggest that men exhibit a higher propensity to adopt AI technologies and report greater confidence in their use (Hechler et al., 2025). This difference in technological orientation may contribute to less critical evaluations of fairness and reduced sensitivity to potential biases in AI systems. These findings indicate that gender shapes both perceptions of AI fairness and attitudes toward AI. Gender differences are therefore likely to influence how fairness perceptions translate into attitudinal outcomes. Accordingly, the following hypothesis is proposed:

H3. 

Gender moderates the relationship between perceived AI fairness and employees’ attitudes toward AI, such that the effect of fairness perceptions on attitudes may differ between men and women.

2.3. Attitudes Toward AI and Creativity

Understanding how employees’ AAI influence workplace behaviors is of critical importance (Savolainen et al., 2026; Koenig, 2025; Schepman & Rodway, 2020, 2023). Among these behaviors, creativity has received particular attention, although existing findings remain inconclusive. The ambivalent nature of AAI contributes to these mixed outcomes, as both enabling and constraining effects on creativity have been identified (Ma et al., 2026; Dong et al., 2025). Positive AAI are typically associated with openness toward AI use, higher levels of trust, and the perception of AI as a resource that supports exploration and innovation. Individuals with such attitudes are more likely to view AI as a tool that facilitates new ways of performing tasks, enhances problem-solving capabilities, and generates efficiency gains as well as broader organizational and individual benefits (Daly et al., 2025; Yoo, 2025; Marocco et al., 2025). In this sense, AI can function as a catalyst for creativity by augmenting human capabilities and enabling the generation of novel ideas. In contrast, negative AAI tend to emerge among individuals who perceive a loss of control or diminished autonomy in their work. These individuals are more likely to view AI as an unreliable or threatening technology that may constrain their ability to think independently and express creativity. Such perceptions can lead to resistance toward AI use and hinder creative engagement by fostering feelings of uncertainty and reduced agency (Daly et al., 2025).

To further explain how AAI influence creative behaviors, this study draws on TAM (Davis, 1989), one of the most widely applied frameworks for explaining how technology use shapes individual outcomes (Savolainen et al., 2026; Koenig, 2025). TAM posits that individuals form evaluations of a technology based on two key determinants: perceived usefulness and perceived ease of use (Davis, 1989). These perceptions shape users’ attitudes toward the technology, which, subsequently, influence their intention to adopt and their engagement with it (Koenig, 2025). Within this framework, when employees perceive AI as both useful and easy to use, they are more likely to develop favorable attitudes and to incorporate it into their daily work practices. Such engagement can enhance creativity by supporting idea generation, improving problem solving skills, and enabling more efficient task execution (Khoso et al., 2026; Ma et al., 2026; Sun et al., 2025; Wang et al., 2025). AI can serve, therefore, as a complementary resource that augments employees’ creative capabilities. In contrast, when AI is perceived as complex and lacking in utility, employees are less likely to adopt or meaningfully engage with it (Koenig, 2025; Tong et al., 2025). Negative perceptions may also reinforce resistance to technology use, further inhibiting creative outcomes. Building on TAM and prior empirical evidence, it is therefore reasonable to expect that more positive AAI are associated with higher levels of creative performance. Accordingly, the following hypothesis is proposed:

H4. 

Employees’ attitudes toward AI positively influence their creativity.

2.4. The Mediating of Attitudes Toward AI

In contemporary organizational settings characterized by rapid and complex technological transformations, particularly those driven by AI, workplace dynamics have become increasingly uncertain and difficult to interpret (Cascio, 2020). Within such environments, intermediate psychological mechanisms assume a central role in explaining how employees process and respond to technological changes. Among these mechanisms, AAI are especially important, as they provide a cognitive and affective lens through which individuals interpret organizational practices, including perceptions of fairness, and translate them into behavioral outcomes. Moreover, the integration of AAI as a mediating variable is grounded in their capacity to reflect individual perceptions within a broader socio-organizational context (Savall & Zardet, 1995). Attitudes are not formed in isolation; they are shaped by organizational norms, cultural expectations, and social interactions surrounding technology use. As such, they capture both individual-level evaluations and contextual influences, making them particularly suitable for explaining variations in AI adoption and use.

As prior research has yet to explicitly examine the relationship between PAF and employee creativity through AAI, the present study builds on established theoretical foundations and related empirical insights. Drawing on SET, PAF can be understood as a critical antecedent of employees’ creative behaviors (H1) (Khoso et al., 2026; Tran Pham, 2023). When employees perceive AI-related practices as fair, transparent, and ethically sound, they are more likely to reciprocate with positive work-related behaviors, including enhanced creativity. In addition, prior research suggests that fairness perceptions influence individuals’ evaluations of AI technologies. When organizational AI practices are perceived as fair and ethical, employees are more likely to recognize the value and benefits of AI, which in turn fosters more positive attitudes toward its use (H2) (Daly et al., 2025; Yoo, 2025; Marocco et al., 2025). These perceptions shape how employees interpret AI systems, affecting their openness to and confidence in engaging in creative tasks with such technologies (Amari et al., 2024). Furthermore, and consistent with TAM, AAI are expected to translate into behavioral outcomes. Employees who hold favorable attitudes toward AI are more likely to adopt and actively integrate these technologies into their work processes. Such engagement can facilitate idea generation, problem-solving, and innovation, thereby enhancing creative performance (H4) (Khoso et al., 2026; Ma et al., 2026; Sun et al., 2025; Wang et al., 2025; Boussioux et al., 2024). These arguments suggest that AAI function as a key psychological mechanism linking PAF influences to employee creativity. Accordingly, the following hypothesis is proposed:

H5. 

Employees’ attitudes toward AI mediate the relationship between perceived AI fairness and their creativity.

Figure 1 below presents the conceptual model proposed in this study. The model illustrates the relationships between PAF, AAI, and employee creativity. Specifically, PAF is expected to exert both a direct effect on creativity and an indirect effect through AAI, which function as a mediating mechanism.

3. Materials and Methods

3.1. Procedure and Data Collection

Prior to the main data collection, a pilot study was conducted to enhance the clarity and the validity of the questionnaire. The pilot involved two management professors and two AI engineer–developers, ensuring that the instrument adequately captured both organizational and technological dimensions. Feedback obtained during this phase led to refinements in item wording, improved contextual relevance, and greater clarity across both language versions of the survey. These adjustments contributed to strengthening the overall validity and reliability of the measurement instrument.

Data were collected via a bilingual (Arabic and English) online questionnaire administered through the Google Forms platform. The questionnaire was distributed primarily through social media channels (e.g., Facebook; WhatsApp) between January and March 2026. To increase participation and reach a broader and more diverse respondent pool, a snowball sampling technique was employed (Amari et al., 2024). Initial participants were encouraged to share the survey link with colleagues within their professional networks, both locally and internationally. This approach facilitated access to a diverse pool of respondents across various organizational contexts and cultural settings.

As shown in

Table 1. Sample Characteristics.

Characteristics	Frequency (N = 214)	Percentage
Gender
Men	103	48.1%
Women	111	51.9%
Age
Less than 29	27	12.6%
Between 30–44	117	54.7%
Between 45–60	66	30.8%
More than 60	4	1.9%
Marital Status
Married	154	72%
Single	50	23.4%
Divorced	9	4.2%
Other	1	0.5%
Education
Undergraduate	12	5.6%
Graduate	74	34.6%
Master and above	128	59.8%
Nationalities
North Africa	101	47.1%
Middle East	84	39.2%
Asia	17	8%
Europe	10	4.7%
North America	2	1%
Occupation field
Education	81	37.9%
Engineering	52	24.3%
Business	38	17.8%
Healthcare	12	5.6%
Hospitality	11	5.1%
Other	20	9.3%

, the final dataset consists of 214 employees and is balanced in terms of gender (approximately 48% men and 52% women). The sample is predominantly mid-career, with most respondents aged between 30 and 44. The sample is multicultural, largely drawn from the MENA region (approximately 47% from North Africa and 39% from the Middle East), with smaller representations from Asia, Europe, and North America. Participants were highly educated, with nearly 60% holding a Master’s degree or higher. This profile reflects a strong readiness to adopt AI technologies (Hechler et al., 2025) and includes employees from advanced sectors such as AI, engineering, and education, often described as “early adopters” (Rodríguez Espíndola et al., 2022). Accordingly, the results should be interpreted with caution, given the specific characteristics of the sample.

3.2. Ethical Considerations

This study was conducted in line with established ethical standards in social science research. Participation was fully voluntary, and respondents were clearly informed that their answers would remain anonymous and would be used solely for research purposes. These measures were designed to ensure confidentiality and transparency throughout the process. In addition, as recommended by Podsakoff et al. (2012), these procedural precautions helped minimize the risk of Common Method Bias (CMB) and reduce the influence of social desirability on participants’ responses.

3.3. Measures

All variables were measured using five-point Likert scales ranging from 1 (strongly disagree) to 5 (strongly agree). Perceived AI Fairness (PAF) as independent variable, was assessed using six items adapted from Ambrose and Schminke (2009) for the AI context. An example item is: “Overall, I feel that my organization treats me fairly when using AI.” Attitudes toward AI (AAI) as mediating variable, was measured using the eight-item short scale developed by Schepman and Rodway (2020). Negatively worded items were reverse-coded to ensure that higher scores consistently indicate more positive attitudes, enhancing clarity and reliability. Creativity (CR), as dependent variable, is measured using a seven-item scale adapted from J. Zhou and George (2001). An illustrative item is: “I actively seek out new creative ideas and inspirations, which I then apply in my work.”

3.4. Analysis Method

This study employs PLS-SEM using Smart PLS 4 to test the proposed hypotheses. PLS-SEM is particularly appropriate for analyzing complex models that incorporate both mediating and moderating relationships, while also accommodating relatively small to medium sample sizes (J. Hair et al., 2020; Henseler et al., 2015). This approach enables the simultaneous assessment of both the measurement and structural models, allowing for a comprehensive examination of the relationships between PAF, AAI, and creativity. The final sample of 214 respondents exceeds the recommended minimum for PLS-SEM, ensuring sufficient statistical power for reliable hypothesis testing (J. Hair et al., 2020).

4. Results

4.1. Measurement Assessment

Confirmatory factor analysis (CFA) was conducted following J. Hair et al. (2020) to evaluate the measurement model and guide PLS-SEM estimation (

Table 2. Reliability and Convergent Validity of Variables.

Constructs	Number of Items Retained	Range of Loadings	$\mathit{\alpha }$	$\mathit{\rho }\mathit{A}$	AVE
AAI	7	0.719–0.834	0.773	0.732	0.611
CR	6	0.796–0.85	0.907	0.914	0.683
PAF	5	0.820–0.925	0.808	0.924	0.545

). Construct reliability was assessed, and items1 with low outer loadings (≤0.7) were removed to improve measurement consistency. All constructs demonstrated satisfactory internal reliability, with Cronbach’s alpha and rho_A values exceeding the recommended threshold of 0.7. Convergent validity was also confirmed, as the average variance extracted (AVE) for each construct was above 0.5. In addition, following Podsakoff et al. (2012), common method variance (CMV) was not deemed a serious concern, as the first factor accounted for less than 50% of the total variance.

Discriminant validity was evaluated using the Heterotrait–Monotrait (HTMT) ratio (J. Hair et al., 2020). As shown in

Table 3. Discriminant Validity Tests.

	AAI	CR	PAF	Square Root of AVE
AAI	1			0.781
CR	0.386	1		0.826
PAF	0.290	0.224	1	0.738

, all HTMT values were below 0.85 and significantly different from 1, indicating that discriminant validity was adequate. Following J. Hair et al. (2020), these results provide strong evidence that the constructs are distinct. The correlation matrix further supports discriminant validity based on the Fornell and Larcker (1981) criterion. Specifically, the square roots of the average variance extracted (AVE), ranging from 0.738 to 0.826, exceed the corresponding inter-construct correlations, confirming that the constructs are satisfactorily distinct from one another.

4.2. Hypotheses Testing

As part of the structural model assessment (Figure 2), multicollinearity was evaluated using the variance inflation factor (VIF). All VIF values were below the recommended threshold VIF < 5. This indicates that multicollinearity was not a concern in the sample (J. Hair et al., 2020).

Next, the proposed hypotheses were examined, with path coefficients and significance tests at the 5% level (critical ratio [CR] > 1.96) summarized in

Table 4. Path coefficients (direct and moderating effects).

	AAI	CR
R2	0.109	0.143
Paths coefficients (CR)
PAF	0.221 (2.535)	0.172 (2.012)
AAI	-	0.299 (4.084)
Gender × PAF	−0.113 (1.395)	-

.

Following J. F. Hair et al. (2017), the coefficient of determination (R2) was used to assess the explanatory power of a structural model, as it measures a model’s in-sample predictive power. As presented in Table 4, the R² values indicate that the explanatory power for creativity, as explained by both AAI and PAF, is greater than the explanatory power for attitudes explained by PAF alone.

Following J. F. Hair et al. (2017), bootstrapping was employed to assess the significance of path coefficients and other estimates in the PLS-SEM analysis. The results of the bootstrapping procedure are presented in Table 4 and

Table 5. The mediation effects (bootstrap with 5000 iterations).

	From	To	Effects	Effects (Bootstrap)	Lower Bound (95%)	Upper Bound (95%)
Specific indirect effects	PAF	CR	0.066	0.078	0.011	0.116

.

The analysis of direct effects indicates that both PAF (β = 0.172, p < 0.05) and AAI (β = 0.299, p < 0.01) have a positive and significant impact on creativity. Thus, hypotheses H1 and H4 are supported. The findings also show that PAF positively affects AAI (β = 0.221, p < 0.05), supporting H2.

Regarding the moderating effect of gender, the interaction between PAF and gender on AAI is not significant (β = −0.113, p > 0.05), and H3 is therefore not supported.

Concerning mediation, PAF significantly affects creativity when AAI are excluded from the model (β = 0.172, p < 0.05). When AAI are included, the effect of PAF on creativity decreases but remains marginally significant (β = 0.066, p < 0.1), indicating partial mediation. Therefore, H5 is partially supported.

5. Discussion

This study investigated the relationship between PAF, employees’ AAI, and individual creativity, while also examining the moderating role of gender in this process. The findings contribute to the growing literature on AI-enabled workplaces by demonstrating that fairness perceptions play a central role in shaping employees’ responses to AI technologies and their engagement in creative behaviors. In the context of rapid digital transformation, where organizations increasingly rely on AI systems to support decision-making and innovation, these findings highlight the importance of integrating ethical and human-centered AI practices into organizational processes. More specifically, this study extends current debates on AI adoption by showing that employees’ reactions to AI are not driven solely by technological efficiency or performance expectations. Rather, perceptions of fairness, transparency, and equitable treatment significantly influence how employees interpret and engage with AI systems. By combining Social Exchange Theory (SET) and the Technology Acceptance Model (TAM), the study provides a more holistic perspective of how organizational and psychological mechanisms jointly shape creativity in AI-enabled environments.

The findings reveal a significant positive relationship between PAF and creativity (H1). This result reinforces previous studies suggesting that fairness perceptions foster innovative and creative work behaviors (Khoso et al., 2026; Tran Pham, 2023). In AI-enabled settings, fairness appears to reduce employees’ uncertainty regarding the use of emerging technologies and creates a work environment characterized by trust and psychological safety. Employees who perceive AI-related practices as fair are therefore more willing to experiment with new approaches, engage in problem-solving activities, and take creative risks (Amari et al., 2024; Amabile & Pratt, 2016; Edmondson, 1999). From the perspective of SET, this finding suggests that employees reciprocate fair organizational treatment by contributing positively to organizational goals through enhanced creativity (Y. Zhou et al., 2026; Yoo, 2025; Amari et al., 2024). When AI systems are implemented transparently and ethically, employees are less likely to perceive them as threatening or controlling mechanisms. Instead, they are more likely to view AI as supportive resources that complement their capabilities and facilitate creative exploration. This interpretation is particularly important because recent debates on AI often emphasize the risks of automation, surveillance, and technological replacement. The present findings instead indicate that fairness perceptions can mitigate these concerns and transform AI into a catalyst for creative engagement.

PAF positively influences employees ‘AAI (H2). This result is consistent with prior studies indicating that fairness and transparency increase trust, acceptance, and openness toward AI technologies (Daly et al., 2025; Marocco et al., 2025; Yoo, 2025). Employees are more likely to develop favorable attitudes when they believe that AI systems operate according to ethical and unbiased principles. This relationship may also be explained by the characteristics of the sample. Most participants were highly educated professionals working in technologically advanced sectors such as engineering, higher education, and business. Such individuals are generally more exposed to digital technologies and may therefore be more capable of recognizing the potential benefits of AI systems (Hechler et al., 2025). Consequently, when fairness is perceived, these employees are more inclined to evaluate AI positively and integrate it into their professional activities. The findings further support the cost–benefit perspective embedded within SET. Employees appear to assess AI technologies by balancing expected benefits, such as improved efficiency and productivity, against potential risks, including bias, surveillance, or job insecurity. When AI systems are perceived as fair and transparent, the perceived benefits outweigh the risks, thereby generating more favorable AAI. These findings reinforce the argument that ethical AI governance is not only a moral requirement but also a strategic organizational resource that shapes employees’ willingness to adopt AI technologies.

It is important to stress again that, contrary to expectations, gender did not significantly moderate the relationship between PAF and AAI (H3), despite a balanced sample distribution. These finding challenges prior assumptions suggesting that men and women differ substantially in their perceptions of AI fairness and adoption (Schepman & Rodway, 2020, 2023; Hechler et al., 2025). One possible explanation is that the increasing democratization of AI technologies has reduced traditional gender disparities in technological engagement. AI systems are becoming more accessible and integrated into everyday professional activities, which may contribute to more homogeneous experiences across genders. Another interpretation relates to the composition of the sample. Participants were predominantly highly educated professionals with advanced technological exposure and experience. Such characteristics may reduce traditional gender-based differences in technology perceptions because both male and female respondents possess relatively similar levels of digital competence and readiness to use AI. This finding may also reflect broader social and organizational transformations associated with digitalization. As AI becomes increasingly embedded within organizational routines, employees may evaluate these technologies more on the basis of practical utility and organizational context rather than gender-based expectations or stereotypes. Consequently, fairness perceptions appear to operate similarly across genders, suggesting that organizational practices and contextual factors may now play a more important role than demographic differences in shaping employees’ AAI.

Another worthwhile finding is the significant positive relationship between AAI and creativity (H4). This result confirms prior research demonstrating that employees who perceive AI as useful and supportive are more likely to engage in creative work behaviors thereby supporting the findings of Khoso et al. (2026), Sun et al. (2025) and Wang et al. (2025). In line with TAM, favorable AAI appear to encourage employees to integrate these technologies into their daily activities, thereby facilitating idea generation, experimentation, and problem-solving. Importantly, the findings suggest that AI should not be viewed merely as an automation tool. Instead, AI can function as a complementary cognitive resource that enhances employees’ creative capacities. Employees with positive AAI may be more willing to collaborate with these systems, use them to explore alternative solutions, and incorporate AI-generated suggestions into their creative processes. The cultural composition of the sample may further explain this relationship. A large proportion of respondents originated from collectivistic cultures within the MENA region, where collaboration, mutual support, and collective achievement are often emphasized (Hofstede et al., 2010). Such cultural orientations may facilitate more positive interpretations of AI technologies (Jecker & Nakazawa, 2022), particularly when these technologies are perceived as tools that support teamwork, efficiency, and shared organizational goals. These contextual characteristics may therefore strengthen the positive relationship between attitudes toward AI and creativity.

This research provides evidence supporting the mediating role of AAI in the relationship between PAF and creativity (H5). The results indicate that PAF not only directly influences creativity but also exerts an indirect influence by shaping employees’ AAI. This finding highlights the importance of psychological and attitudinal mechanisms in explaining how fairness perceptions translate into behavioral outcomes. More specifically, the findings suggest that fair AI practices reduce anxiety, uncertainty, and resistance toward AI systems, thereby increasing employees’ confidence in using these technologies creatively. Employees who perceive AI systems as fair are more likely to develop positive attitudes, which subsequently encourage experimentation, creative problem-solving, and innovative thinking (Ma et al., 2026; Sun et al., 2025; Dong et al., 2025). However, the mediation effect was only partial, indicating that PAF continues to exert a substantial direct influence on creativity beyond employees’ AAI. This suggests that fairness itself constitutes an important organizational resource capable of directly stimulating creativity, regardless of employees’ attitudinal evaluations. One explanation may be that fair organizational environments promote psychological safety, trust, and engagement independently of technology perceptions. Additionally, the relatively weaker indirect effect through AAI may be associated with the profile of the respondents. Because participants demonstrated high levels of education, digital competence, and technological familiarity, their engagement with AI may depend less on attitudinal mechanisms than would be the case for less technologically experienced populations. In other words, employees with advanced technological readiness may engage creatively with AI even when attitudinal influences are less pronounced.

Overall, the findings contribute to a more nuanced understanding of AI-enabled creativity by demonstrating that organizational fairness, employee attitudes, and contextual characteristics jointly shape how individuals engage with AI technologies. The results further emphasize that the successful implementation of AI in organizations depends not only on technological sophistication but also on ethical governance, transparent practices, and employees’ psychological readiness to collaborate with intelligent systems.

5.1. Theoretical Implications

The findings underscore the importance of examining both direct and indirect pathways in studying AI adoption within the organizations, thereby offering a more nuanced understanding of how professionals engage in creative tasks involving emerging technologies across cross-cultural context. This perspective is especially relevant for future research conducted in diverse cultural and economic settings, as it highlights the need for flexible theoretical frameworks that accommodate context-specific dynamics.

First, this study provides original empirical evidence on the importance of inclusive organizational practices (e.g., PAF) in fostering positive attitudes and behaviors in AI-augmented contexts, contributing to the growing body of literature on AI ethics (Yoo, 2025; Köchling et al., 2025). More precisely, findings from this study extend prior work (Daly et al., 2025) by showing that a fair organizational climate, combined with effective human–AI collaboration, enhances employee creative performance.

Furthermore, this research advances justice research by integrating employees’ AAI as a boundary condition in the link between PAF and creativity. It appears that these attitudes shape sensemaking processes related to technology use (Yoo, 2025), influencing how employees interpret organizational practices and translate perceptions of fairness into creative behaviors. This suggests that the deployment of AI should be accompanied by fair and transparent practices to foster employees’ creativity within organizations.

Additionally, by examining the moderating role of gender, this study enriches our understanding of how male and female employees respond to AI technologies. Given the limited empirical evidence on gender-based differences in AI-related perceptions (Schepman & Rodway, 2020, 2023), the findings provide nuanced insights, suggesting that gender differences are less salient within AI-enabled organizational settings. This result points to a convergence in how men and women engage with AI technologies, indicating that female employees may be equally empowered as their male counterparts in navigating technological change. Hence, the results of this study perhaps challenge traditional assumptions regarding gender-based disparities in responses to emerging technologies.

Finally, this study proposes an integrative theoretical framework that combines TAM and SET to provide a more comprehensive explanation of the dynamics between PAF, AAI, and creativity. While TAM traditionally focuses on cognitive and technological determinants of adoption, the inclusion of SET allows for the incorporation of relational and social dimensions, particularly the role of PAF within organizational contexts. Such an approach enhances the explanatory power of existing models and offers a more holistic understanding of employees’ attitudes and behaviors in AI organizational settings.

5.2. Managerial Implications

From a managerial perspective, the findings highlight the critical need for an ethically grounded approach to the development and implementation of AI in creative tasks. Organizations should ensure that AI applications are deployed in a responsible and transparent manner that supports, rather than constrains, innovation. In this regard, organizations should invest in tailored training programs, awareness initiatives, and skill development workshops aimed at clarifying AI functionalities, reducing misconceptions, and enhancing employees’ competence and readiness to use AI.

Beyond formal training, managers are encouraged to implement participatory approaches such as co-design initiatives that involve employees in the process of AI integration. Such practices can strengthen the PAF, increase engagement, and contribute to a more human-centered implementation of AI within organizational settings.

Finally, maintaining open communication channels is essential. Managers should encourage employee voice through regular meetings, structured feedback mechanisms, and idea-sharing platforms. These initiatives enable employees to express concerns, seek clarification, and contribute suggestions, thereby facilitating smoother and more effective adoption of AI in creative work processes.

5.3. Limitations and Future Directions

Although the findings highlight the relevance and contributions of this study, some limitations should be acknowledged, which also provide directions for future research. First, the use of an online questionnaire did not allow participants to directly experience the situations under investigation, which may have influenced their responses. Second, the study relies on a sample drawn primarily, but not exclusively, from the MENA region, which is characterized by specific cultural features. This may introduce issues related to cultural heterogeneity and potentially limit the generalizability of the findings. Third, the cross-sectional design restricts the ability to establish causal relationships. Future research could address these limitations by adopting longitudinal designs or scenario-based causal chain approaches (Köchling et al., 2025). In addition, examining employee responses within real organizational settings would provide a more in-depth understanding of how perceptions are formed and enacted. Furthermore, conducting a multi-group analysis would be advisable to compare relationships across different cultural clusters and mitigate potential heterogeneity bias. Finally, this study considers AI in a general sense. Future research would benefit from focusing on specific AI applications, as perceived value and adoption may differ across organizational contexts and individual characteristics (Papagiannidis et al., 2023).

6. Conclusions

In the context of AI democratization across sectors and among individuals, organizations are striving to ensure their sustainability by fostering innovative behaviors. This study contributes to this objective by emphasizing the importance of inclusive practices and demonstrating how ethical and equitable AI implementation can support authentic and sustainable creativity. More importantly, a key finding of this study suggests a potential shift toward a more egalitarian and inclusive organizational landscape in the context of digital transformation, in which traditional differences, particularly those related to gender, tend to become less salient. The findings also indicate that PAF plays a critical role in shaping positive attitudes and engagement, thereby contributing to both individual and organizational performance. By combining fair AI practices with the active management of employee attitudes toward AI, organizations can create a learning environment in which individuals feel confident, supported, and motivated to use AI in ways that enhance creativity and innovation.