Curious Minds, Open Offices: What Shapes How Workers Really Feel About AI

As artificial intelligence spreads through workplaces at a pace few predicted, one question has moved from the seminar room to the boardroom: why do some employees embrace AI with enthusiasm while others greet it with suspicion? A new study published in Current Psychology offers a carefully constructed answer, suggesting that the road to positive attitudes toward AI runs through two very different territories — the learning habits of individual workers and the innovation climate of the organizations that employ them. The research, based on survey data from 423 corporate employees in China, maps how self-directed learning, organizational context, and personality interact to shape how people evaluate artificial intelligence in their working lives.

The study, conducted by Wu Zhangxun of the Seoul School of Integrated Sciences and Technologies, draws on two well-established theoretical foundations: the five-dimensional curiosity framework and technology acceptance theory. Technology acceptance theory, rooted in the influential work of Fred Davis and later expanded into unified models of technology adoption, holds that people’s willingness to embrace new technologies depends heavily on their perceptions of usefulness and ease of use. Attitudes toward technology, in this view, are not fixed dispositions but evaluations that can be shaped by experience, context, and motivation. The five-dimensional curiosity framework, developed by Todd Kashdan and colleagues, breaks curiosity into distinct facets — including joyous exploration, deprivation sensitivity, stress tolerance, social curiosity, and thrill seeking — recognizing that curiosity is not a single switch but a spectrum of motivational tendencies.

At the heart of the study lies a concept with a long pedigree in adult education: self-directed learning. First formalized by Malcolm Knowles in the 1970s, self-directed learning describes the process by which individuals take initiative in diagnosing their own learning needs, setting goals, identifying resources, and evaluating outcomes, with or without help from others. In an era when the half-life of professional skills keeps shrinking, self-directed learning has been described as a fundamental competence for a rapidly changing world. The study’s central hypothesis was straightforward: employees who habitually drive their own learning should hold more positive general attitudes toward artificial intelligence, because learning-oriented people are more likely to encounter AI on its own terms, understand its capabilities, and integrate it into their sense of professional identity rather than experiencing it as a threat.

To test this idea, the researcher built a moderated mediation model — a statistical architecture that allows a single study to answer several questions at once. The first question is whether self-directed learning predicts general attitudes toward artificial intelligence. The second concerns mechanism: what carries that influence from one to the other? Here the study nominated innovation climate, the shared perception among employees that their organization encourages new ideas, tolerates experimentation, and rewards creative effort. The third question concerns boundary conditions: under what personal circumstances does the relationship strengthen or weaken? For that role, the study enlisted five-dimensional curiosity as a moderator, testing whether curious employees convert their learning into more favorable AI attitudes more readily than their less curious peers.

The data told a coherent story. Self-directed learning was positively associated with employees’ general attitudes toward artificial intelligence, confirming the basic prediction that people who own their learning tend to view AI more favorably. More intriguingly, innovation climate mediated this relationship. Employees who engaged in self-directed learning perceived a stronger innovation climate around them, and that perception, in turn, was linked to more positive attitudes toward AI. The interpretation is subtle but important: learning-oriented employees do not simply absorb new technology in isolation. They appear to read their environment, and when their own proactive learning resonates with an organization that visibly values innovation, the combination translates into a more welcoming stance toward artificial intelligence.

The moderating results added a second, distinct layer to the picture. Five-dimensional curiosity moderated the direct link between self-directed learning and attitudes toward AI, but it did not significantly moderate the relationship between self-directed learning and innovation climate. In practical terms, curiosity changes how employees personally evaluate AI once they have learned about it, but it does not change how they perceive their organization’s innovativeness. The author concludes that innovation climate and employee curiosity operate through distinct mechanisms — one organizational and contextual, the other individual and motivational — and that both matter for how employees evaluate AI within organizational settings. This separation matters theoretically, because it suggests that the environmental pathway and the personality pathway are parallel rather than intertwined, each contributing independently to the formation of technology attitudes.

The study’s methodological choices deserve attention. The survey instruments were grounded in validated measures, including established scales for self-directed learning and general attitudes toward artificial intelligence, the latter a construct that has been formalized in recent years through dedicated psychometric instruments designed to capture attitudes toward AI as a general phenomenon rather than a specific application. The analysis followed modern conventions for testing mediation and moderation, drawing on statistical frameworks for conditional process models that have become standard in organizational psychology. The researcher also took steps associated with controlling common method bias, a well-known hazard in single-source survey research, and the work was conducted in compliance with the ethical principles of the Belmont Report, with written informed consent obtained from all participants for the publication of anonymized data.

Context matters for interpreting these findings. China has made artificial intelligence a national priority since the State Council’s 2017 New Generation Artificial Intelligence Development Plan, and state-owned enterprises in particular have been pushed to modernize through digital transformation initiatives. That policy backdrop makes Chinese corporate employees a revealing population for studying AI attitudes: they sit at the intersection of rapid technological deployment, organizational reform, and intense public discussion about the future of work. Attitudes toward AI in such an environment are not abstract opinions; they shape whether training programs succeed, whether AI tools are actually used once purchased, and whether employees experience technological change as opportunity or displacement.

For organizations, the practical implications are tangible. If innovation climate carries part of the effect of self-directed learning on AI attitudes, then companies cannot simply exhort employees to be curious about AI and expect results. They need to build visible, credible environments in which experimentation is safe, new ideas circulate, and learning is resourced — conditions long associated in the organizational literature with psychological safety and creative work environments. At the same time, because curiosity operates on the individual pathway, recruitment and development practices that identify and cultivate curious employees may pay dividends specifically in technology adoption. The two levers are complementary: one shapes the stage, the other shapes the actor.

The study also carries the usual caveats of cross-sectional survey research. Measuring learning, climate, curiosity, and attitudes at a single point in time means the causal ordering, however plausible theoretically, is inferred rather than observed — a limitation that methodologists have long flagged for mediation models of this kind. The sample, drawn from corporate employees in China, invites replication in other national and cultural contexts before the findings are generalized. And attitudes toward AI are a moving target: as generative tools reshape daily work, the meaning of a positive AI attitude itself may shift. Still, the study’s core contribution stands. It disentangles two forces — a contextual one and a personal one — that are often blurred together in discussions of technological change, and it suggests that winning hearts and minds for AI in the workplace requires attention not only to what employees learn, but to where they learn it and who they are when they do.

Subject of Research: The relationship between corporate employees' self-directed learning and their attitudes toward artificial intelligence, mediated by innovation climate and moderated by five-dimensional curiosity.

Article Title: Corporate employees’ self-directed learning and attitudes toward artificial intelligence: the mediating role of innovation climate and the moderating role of five-dimensional curiosity

Article References: Zhangxun, W. (2026). Corporate employees’ self-directed learning and attitudes toward artificial intelligence: the mediating role of innovation climate and the moderating role of five-dimensional curiosity. Current Psychology, 45(17), Article 1493. https://doi.org/10.1007/s12144-026-10045-4

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10045-4

Keywords: artificial intelligence, self-directed learning, innovation climate, five-dimensional curiosity, technology acceptance, corporate employees, organizational psychology, attitudes toward AI, workplace learning, moderated mediation, Current Psychology, China