Most workers say AI makes them faster — so why do they hide that from their own coworkers and bosses?

This explores why people who say AI makes them faster at work still keep that quiet around colleagues and managers, and what that silence costs. The basic pattern is well documented. In Anthropic's interview study of 1,250 people, 86% of general workers and 97% of creatives said AI saved them time, yet roughly 70% mentioned the stigma of using AI at work. They pointed to workplace stigma, worries about their professional identity and fear of being replaced Why do workers hide productivity gains from AI use?. Controlled experiments show the fear is something people predict ahead of time. Across four studies with more than 4,400 participants, AI users expected to be judged less competent and less diligent, and they became less willing to tell managers and colleagues about the tool Do people fear judgment when they use AI at work?.

Fear of judgment is only part of the story, though. Interviews with knowledge workers suggest that removing signs of AI from their work is also a way to show expertise. Polished output with no visible AI help reads as a sign of skill, so hiding it is self-promotion as well as self-protection Why do knowledge workers hide signs of using GenAI?. Colleagues' suspicion isn't baseless either. In a BetterUp Labs survey, more than half of desk workers admitted sending AI-generated 'workslop' (content that looks finished but says little), and recipients reported losing hours a month to it and trusting the senders less How much do desk workers actually suffer from AI workslop?. Some of the stigma people expect is a reasonable response to low-effort AI output.

The less obvious cost lands on the organization. When people hide how they use AI, informal learning between colleagues dries up. Nobody sees the useful prompt or the workflow that actually works Why do knowledge workers hide signs of using GenAI?. That may help explain a puzzle in McKinsey's data. 80% of respondents report personal AI productivity gains, but only 37% see company profits improve, and the few firms that do capture value redesign workflows openly rather than leaving AI to individual habits Why do individual AI gains not translate to business results?. Executives also tend to see bigger gains than the measurements show Do AI productivity gains feel larger than they actually measure?, and quiet, private use makes that gap harder to check. Microsoft's argument that the next step is team-level rather than individual productivity Can AI boost how teams work together? assumes the kind of visibility that concealment prevents.

There are signs this can change. Writers sharing a document preferred editors that showed more of each other's prompting, because seeing when and how a partner used AI helped them follow the partner's thinking and check the text. Some still found full visibility intrusive Do writers want to see each other's AI prompts in shared editors?. Research on disclosing AI identity finds that people's first reaction is biased against AI, but the bias reverses once they repeatedly see good results Does revealing AI identity help or hurt user trust?. Taken together, these suggest that disclosure alone mostly invites judgment. What reduces stigma is visible results over time.

One cross-domain parallel: Hoel argues that AI is pushing researchers to keep unfinished ideas secret, because AI can now finish someone's partly shared idea before they do Does AI scooping force researchers to hide work in progress?. In both offices and research, AI rewards secrecy for the individual while reducing the open sharing that makes groups smarter. Note that this corpus documents the hiding and its costs well but has little on interventions that have actually worked to reverse it.