CEOs say AI depends on adoption. They estimate a quarter of staff use it.

Of the chief executives in IBM's 2026 CEO Study, 83 percent said AI success depends more on people's adoption than on technology. By the same leaders' estimate, only 25 percent of their workforce uses AI regularly. Yet 86 percent believe their employees have the skills.

The IBM Institute for Business Value and Oxford Economics surveyed 2,000 CEOs and equivalent senior leaders in 33 geographies and 21 industries between February and April 2026. The 25 percent is the CEOs' estimate for their total workforce, not measured usage. IBM, which published the study, sells AI software and consulting.

Read together, the answers say something about where CEOs think the problem lies. If employees have the skills but only a quarter use AI regularly, then on the CEOs' own account something other than skills is holding usage back.

A chief executive who believes adoption decides AI's success should not have to estimate it.

Other surveys ask other people

Other 2026 surveys report higher use and wider benefits. They asked different people different questions, and their figures should not be set against IBM's as if they measured the same thing.

BCG's 2026 AI at Work survey of 11,749 workers in 14 markets found that 74 percent of frontline employees use AI regularly. BCG defines frontline employees as individual white-collar employees without managerial responsibilities, not deskless or shop-floor workers. The figure comes from workers describing their own use, in a narrower group than the total workforce IBM's CEOs were estimating for. The two numbers are not comparable, and the distance between them does not show that CEOs underestimate use.

BCG also found that 42 percent of frontline employees who use AI regularly save at least a full workday a week, by their own estimate. Yet 66 percent say they get limited or no guidance on what to do with that time. BCG's release does not state the base for the 66 percent; it appears to be regular frontline users.

McKinsey's 2026 State of AI survey of 1,719 participants in 97 nations is a third population again. It found that 80 percent of respondents say AI has improved their individual productivity, while only 37 percent attribute any EBIT impact to AI, essentially unchanged from 2025. Both are self-reported perceptions, not audited financials.

Measure adoption instead of estimating it

Every figure in the table comes from what people say, not from what systems record. A chief executive who believes adoption decides AI's success should not have to estimate it.

• Define regular use, then report it monthly from system data by function and level. Pair it with an anonymous survey to catch use of personal tools, which system data will not show.
• Ask regular users what happens to the time AI saves them, and decide team by team what those hours are for. BCG's 66 percent figure suggests many users are left to work that out alone.
• Test the skills assumption. Most of IBM's CEOs believe their employees have the skills; a short practical assessment in one function would show whether that holds.
• Report one business measure next to usage. McKinsey's respondents report personal productivity gains far more often than any EBIT impact, so usage and value need separate measures.