What is shadow AI?
Shadow AI is the use of AI tools at work that an employer has not provided or approved, and it is the norm rather than the exception: 78% of people who use generative AI at work bring their own AI tools, according to Microsoft and LinkedIn's 2024 Work Trend Index. Microsoft calls this BYOAI, for bring your own AI. It was even more common at small and medium-sized companies (80%).
KPMG and the University of Melbourne found the same pattern in a larger 2025 study of 47 countries. Among employees who use AI at work, 70% access free, publicly available AI tools, 42% use tools their employer provides and 18% pay for public tools themselves. General-purpose chatbots such as ChatGPT, Copilot and Claude are the most common tools, used by 73%.
Which generation uses shadow AI the most?
Gen Z does, but every generation does it. Microsoft found that 85% of Gen Z knowledge workers who use AI at work bring their own tools, compared with 78% of millennials, 76% of Gen X and 73% of boomers and older. The spread from youngest to oldest is only 12 points.
The bigger generational gap is in who uses AI at all. In CybSafe and the National Cybersecurity Alliance's 2024 survey of more than 7,000 people in seven countries, 72% of Gen Z used generative AI at work, at home or both, against 62% of millennials, 38% of Gen X, 15% of baby boomers and 7% of the Silent Generation. In KPMG's study, 72% of employees aged 18 to 34 used AI at work, compared with 40% of those 55 and older.
How many employees share sensitive data with AI tools?
Almost half: 48% of employees in KPMG's study said they had uploaded company information, such as financial, sales or customer information, into a public AI tool, and 34% said they had done so sometimes, often or very often. A similar share (49%) had uploaded copyrighted material or intellectual property to a generative AI tool.
CybSafe and the National Cybersecurity Alliance found that 43% of respondents in their 2025 survey admitted sharing sensitive workplace information with AI tools without their employer's knowledge, up from 38% of employed participants a year earlier. ChatGPT was the most used tool in that survey, at 77% adoption. In the same 2025 report, 58% of AI users said they had received no training on the security or privacy risks.
Are younger workers more likely to misuse AI at work?
Yes. In KPMG's study, 65% of employees under 35 reported complacent or inappropriate AI use at work, compared with 53% of those aged 35 to 54 and 41% of those 55 and older. KPMG says the difference holds even after accounting for how often each group uses AI, and that employees with AI training and higher incomes also report more of these behaviors.
Policy is not keeping up. Only 40% of employees said their workplace has a policy or guidance on generative AI use, and 56% said they had used AI tools at work without knowing whether it was allowed. KPMG found that uploading sensitive information was most common among employees whose organization had banned generative AI (67%).
Why do employees hide their AI use?
Many worry about how it looks. In the 2024 Work Trend Index, 52% of people who use AI at work said they were reluctant to admit using it for their most important tasks, and 53% worried that using it on important work makes them look replaceable. In KPMG's 2025 study, 57% of employees said they had used AI in non-transparent ways, such as avoiding revealing when they used AI or presenting AI-generated content as their own.
Which age group uses AI chatbots at work the most?
Workers under 50 do. In Pew Research Center's February 2026 survey, 41% of employed adults aged 18 to 29 and 43% of those aged 30 to 49 said they use AI chatbots for tasks at work, compared with 29% of those aged 50 to 64 and 18% of those 65 and older. A year earlier, 38% of employed adults under 30 said they had used ChatGPT on the job, against 30% of those 30 to 49 and 18% of those 50 and older.
ChatGPT dominates in every age group. In 2026, 61% of adults under 30 said they use it, compared with 19% of those 65 and older, a 42-point gap. Gemini (30% vs. 11%) and Claude (10% vs. 1%) follow the same pattern on a smaller scale.
Methodology & caveats
The surveys measure different things and are not interchangeable. Microsoft and LinkedIn surveyed 31,000 full-time employed or self-employed knowledge workers in 31 markets online between 15 February and 28 March 2024 (fieldwork by Edelman Data & Intelligence); its generation figures are shares of people who use generative AI at work, not of all workers, and "bring their own AI tools" means using tools not provided by their organization. KPMG and the University of Melbourne surveyed 48,340 people in 47 countries, including 32,352 employees, between November 2024 and January 2025; its age groups are 18–34, 35–54 and 55+. Pew's figures are for U.S. adults (5,119 surveyed 17–23 February 2026; 5,123 surveyed 24 February–2 March 2025). CybSafe and the National Cybersecurity Alliance polled more than 7,000 people in seven countries in 2024 and more than 6,500 in seven countries in 2025.
KPMG values were read from the data labels of its published charts (Figures 40, 44, 52 and 53) and from the report text; its "complacent use" measure combines several behaviors and counts anyone who answered "rarely" or more often. The 2025 Oh, Behave! figure of 43% is reported by the National Cybersecurity Alliance without a stated base, while the 2024 figure of 38% is a share of employed participants, so the rise is indicative. Generation and age labels follow each publisher. Periods differ, from early 2024 (Microsoft) to February 2026 (Pew). The 42-point ChatGPT gap, the "more than twice" comparison (41% vs. 18%) and the "ever" shares in Fig. 5 (100% minus "never") are LLM.co calculations from published figures. The lead for this graphic was a Visual Capitalist chart of Microsoft's generation data; every figure here was checked against the original publisher.
Sources
- [1]Microsoft & LinkedIn. 2024 Work Trend Index Annual Report: AI at Work Is Here. Now Comes the Hard Part (8 May 2024)https://www.microsoft.com/en-us/worklab/work-trend-index/ai-at-work-is-here-now-comes-the-hard-part
- [2]KPMG & University of Melbourne. Trust, attitudes and use of artificial intelligence: A global study 2025 (global report)https://kpmg.com/content/dam/kpmgsites/xx/pdf/2025/05/trust-attitudes-and-use-of-ai-global-report.pdf
- [3]KPMG. Trust, attitudes and use of AI: A global study 2025 (study overview)https://kpmg.com/xx/en/our-insights/ai-and-technology/trust-attitudes-and-use-of-ai.html
- [4]National Cybersecurity Alliance & CybSafe. STUDY: 65% Now Use AI, but Majority Remain Untrained on Risks (30 Sep 2025)https://www.staysafeonline.org/press/study-65-now-use-ai-but-majority-remain-untrained-on-risks
- [5]CybSafe & National Cybersecurity Alliance. The GenAI story isn't just about technology (Oh, Behave! 2024–2025 findings, 21 Oct 2024)https://cybsafe.com/blog/the-genai-story-isnt-just-about-technology
- [6]Pew Research Center. How Americans' opinions and use of AI differ by age (Americans and AI 2026, 17 Jun 2026)https://www.pewresearch.org/internet/2026/06/11/how-opinions-and-use-of-ai-differ-by-age/
- [7]Pew Research Center. 34% of U.S. adults have used ChatGPT, about double the share in 2023 (25 Jun 2025)https://www.pewresearch.org/short-reads/2025/06/25/34-of-us-adults-have-used-chatgpt-about-double-the-share-in-2023/











