Why Workers Are Afraid to Admit They Use AI
AI tools are embedded in work whether we acknowledge them or not. The question is not whether people use AI—it’s whether they can use it with dignity. The shame surrounding AI usage is not inevitable. It reflects uncertainty, misaligned incentives, and cultural expectations that no longer match how work is done.
Photo by Yan Krukau
Across workplaces today, artificial intelligence has quietly become embedded in daily routines. Employees use it to refine reports, draft correspondence, and analyse data with a speed that traditional methods cannot match. However, despite its growing use, many hesitate to acknowledge this reliance openly. The reluctance stems less from technical limitations than from cultural perceptions—concerns about reputation, fears of diminished credibility, and uncertainty over how colleagues or managers will interpret such disclosure. What emerges is not a debate about technology alone, but a broader question of how organisations define effort, competence, and value in an era increasingly shaped by AI.
Several recent studies make it clear that reluctance to disclose AI use isn’t harmless diffidence—it springs from real concerns. A global survey by Microsoft and LinkedIn found that more than half of knowledge workers hide their use of AI for core tasks, and over half fear being viewed as less competent or even replaceable through automation.
In work led by researchers from Duke University, employees using tools like ChatGPT, Claude, or Gemini were judged more harshly by peers and hiring managers—seen as less driven or less authentic when they sought help from AI. These judgments held even when the final results were substantially improved by AI.
Such findings suggest a paradox: the same tools that can enhance productivity are treated as liabilities because their use is perceived as cutting corners. Workers often anticipate negative perceptions, and those expectations sometimes prove correct. In short, AI shame is not just a feeling—it is a professional risk.
The roots of shame are many: fear of being seen as lazy is frequent. If you rely on AI, the argument goes, you are not doing the work yourself. Colleagues or managers may question your intellectual rigor or creativity. Will you be seen as less dedicated? These anxieties are not idle: several studies show that performance-boosting AI use can come with reputational costs.
Another concern is job security. Employees worry that openly using AI may invite scrutiny or lead managers to assume the tasks could be automated entirely. In a climate of cost cutting, layoffs, and reorganisations, being perceived as replaceable can feel dangerous.
However, another factor is the lack of clear policy. Where employers have no guidance or formal position on AI usage, ambiguity reigns. Workers operating in “shadow AI” mode—using AI tools without approval—face the double bind of needing to improve efficiency while avoiding blame for policy violations or errors. Studies show that a large percentage of workers admit to covert AI use, often because they don’t trust that disclosing the truth will lead to support rather than censure.
Cultural factors also play a role. Many professional environments—especially those with rigid hierarchies—still value visible effort, long hours, and traditional indicators of merit (handwritten notes, done-by-hand work). AI, by contrast, can make craftsmanship less visible. Workers who grew up with the idea that “good work” is sweat, not delegation or assistance, may feel that admitting to using AI is confessing to a moral failing.
Operating covertly comes at a price. When people hide their use of AI, it constrains innovation and collaboration. Colleagues cannot learn from each other’s tools or tips, and leadership remains unaware of the potential efficiencies, which limits investment in training or infrastructure.
Secrecy also imposes personal stress. Constantly concealing methods, worrying about being “found out,” editing work manually to mask AI assistance—all of this expends mental energy. People report lower job satisfaction when they feel they cannot speak openly about how they accomplish their work.
Moreover, when reputational penalties exist (real or perceived), they affect promotion, compensation, and professional opportunities. Studies show that employees who disclose AI use may be judged as less motivated or less competent—perceptions that could skew hiring, performance reviews, or responsibility allocation. The penalty exists even when outcomes improve.
If organizations are to reap the benefits of AI—and avoid the stealth burden on workers—they must shift culture and policy.
First, leadership must model openness. Companies that encourage and reward responsible AI use (versus punishing mistakes) reduce stigma. If senior staff acknowledge their own usage—and clarify expectations—others will follow.
Second, formal policies matter. Institutions need clear guidelines on what is acceptable use of AI for tasks, how to credit AI-assisted work, how to handle errors, and how to protect employee data and intellectual property. Without clarity, people will continue hiding in ambiguity.
Third, training and education must go beyond “how to use” to include “how to use well and ethically.” When workers understand the limitations, risks, and best practices of AI, they are less likely to be judged harshly and more likely to feel confident in disclosure.
Fourth, shift performance metrics. Reward output, insight, and impact, not just visible process or long hours. If companies evaluate success by what gets done—not how it was produced—then using AI effectively becomes a strength, not a stigma.
About The Author
