A new study from Duke University reveals a surprising downside for employees using artificial intelligence tools in the workplace: social stigma. While AI tools like ChatGPT and Gemini are praised for boosting productivity, the research warns that coworkers and managers often judge AI users as lazy, less competent, and dependent.
The study, published in the Proceedings of the National Academy of Sciences, included over 4,400 participants across four experiments. Researchers from Duke’s Fuqua School of Business found that employees who used AI to complete tasks were consistently rated lower in diligence, independence, and self-confidence than peers who used traditional tools—or no help at all.
The stigma was consistent across demographic groups, suggesting a broad cultural skepticism toward AI use. This perception led many workers to hide their AI use, fearing career damage or backlash, especially in companies with explicit bans on AI-generated work.
Real-world consequences were evident in a hiring simulation. Managers who didn’t use AI themselves were significantly less likely to hire job candidates who did. In contrast, AI-literate managers showed a preference for fellow users, highlighting the role of personal familiarity in shaping attitudes.
Researchers discovered the main driver of this bias is the perception of laziness. However, when AI was shown to clearly improve task outcomes, the negative evaluations dropped sharply. This finding suggests that transparency and demonstrated utility could help shift attitudes.
For employers eager to integrate AI tools, this presents a serious challenge. Adoption may stall if social penalties persist, and organizations could see division between users and non-users. The hidden cost of AI in the workplace may be less about ethics or capability—and more about human judgment.
