Copilot in the wings: When do employees reveal the use of AI?

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

As generative AI becomes a part of the workplace, employees face a critical decision: whether or not to disclose their use of it. Although the drivers of AI adoption are well-documented, the reasons for disclosure remain largely unexplored. Based on data from employees in the United States (N = 1,011) and using a configurational approach, we uncover configurations of individual-, workplace-, task-, and tool-related factors to explain AI disclosure at work. Our findings reveal three alternative, consistently sufficient configurations for AI disclosure that differ in their composition, but that are conceivable as equifinal pathways to AI disclosure. Knowledge of these patterns of factors contributes to a better understanding of complementarity effects among factors in predicting AI disclosure, and reveals trade-offs that employees make when they disclose the use of AI at work. The results also offer insights for designing AI transparency rules in the workplace in an era of increasing human-AI interaction.