What the Data Actually Shows About AI Use at Work Right Now
Ask five surveys how many people use AI at work and you'll get five different numbers -- here's why, and what to actually take from them.
How many people are actually using AI at work depends a lot on how you ask the question. Pew Research Center's October 2025 survey put it at 21% of U.S. workers using AI on the job -- up from 16% a year earlier -- with only about 10% using it daily. A larger SHRM survey of nearly 6,000 U.S. workers in early 2026 found a higher number, 41%, when the question included any work use at all, and split evenly on whether their employer had formally rolled out AI tools (47% yes, 47% no).
The gap between those numbers is the real story: adoption is uneven and accelerating, not universal and settled. It also varies sharply by role -- software engineers report the highest daily use of any function surveyed, at 71%, followed by marketing (63%) and IT/information security (55%). Plenty of other roles are still in single digits.
There's also a sizable "shadow AI" economy: more than 90% of workers in some surveys say they use personal AI tools for job tasks even when their employer hasn't sanctioned any -- often getting more value out of those unofficial tools than the ones their company officially provides. If your own workplace feels behind or ahead of "everyone," both are true somewhere.
Figures from Pew Research Center (October 2025) and SHRM's 2026 workplace AI survey of 5,875 U.S. workers.
Editorial note: Knowledge Hub pieces are member-submitted and reviewed before publication. They are not scraped from third-party sites. Research datasets live on the AI Job Risk Index.
Related
The 8-Step Framework for Building an AI Agent (Even a Simple One)
Most "AI agent" projects fail for a boring reason: someone picked a model and started prompting before deciding what the thing was actually supposed to do.
How to Read Your AI Exposure Score Without Misreading It
An exposure score is a map of which tasks are more assistable — not a pink slip, not a personality verdict, and not a guarantee about next year's headcount.
LLM vs. RAG vs. AI Agent vs. Agentic AI: What Each Term Actually Means
The buzzwords get used interchangeably in almost every AI conversation now, but they describe four genuinely different things stacked on top of each other, not four names for the same idea.