Overcoming Barriers to AI Disclosure: The Role of Granularity and Sociocultural Enablers in Reducing Perceived Stigmas

Despite growing use of generative AI in the workplace, common stigmas depict AI users as lazy or incompetent and their work as low-quality or untrustworthy. Concerns about these stigmas deter people from disclosing their use of AI, resulting in widespread concealment and harms arising from this lack of transparency. Through a survey of workers who use AI (N=162), we explore whether disclosures that offer additional granularity beyond simple acknowledgments of AI usage can mitigate reservations around disclosure. We found that disclosures indicating compliance with AI policies, how AI contributed, and presence of human review are effective in increasing users’ comfort with disclosure and reducing perceived stigmas. However, granular disclosures are insufficient to meaningfully change disclosure decisions on their own. We contextualize these findings in an ecosystem of sociocultural enablers and barriers identified from participants’ disclosure experiences and share implications for supporting AI disclosure in the workplace.