90% of executives believe AI has not boosted productivity at their companies

Ninety percent of executives believe AI has not boosted productivity at their companies, according to a National Bureau of Economic Research working paper cited by University of Pittsburgh professor Mark Ma in a Fortune article published August 22. Ma's research, analyzing millions of employee reviews and thousands of corporate reports, found that AI-attributed layoffs correlate with negative employee sentiment and no significant stock market gains, suggesting AI is used as cover for cost-cutting rather than genuine efficiency improvements.

Ninety percent of executives admit AI hasn’t actually made their companies more productive. Read that number again, then consider that many of those same executives are still citing AI as the reason for laying people off. Something in that equation doesn’t add up – and new research says it’s not a coincidence, it’s cause and effect. The Research The findings come from Mark Ma, a professor of business administration at the University of Pittsburgh, writing in an article published by Fortune in partnership with The Conversation on August 22. Ma cited a National Bureau of Economic Research working paper from the Atlanta Federal Reserve finding that roughly 90% of executives believe AI has not yet boosted productivity at their companies, and noted that other research points to remote work, rather than AI, as a more likely driver of the productivity gains companies have seen since 2021, according to reporting cited from the New York Times and Stanford economist Nicholas Bloom. Ma and his colleagues analyzed millions of job-satisfaction reviews alongside thousands of corporate financial reports, and hundreds of AI investment and layoff announcements made by U.S. public companies over the past five years, according to the Fortune article. They found that as the frequency of AI investment announcements rose, so did announcements of AI-attributed layoffs, a pattern Ma described as reflecting a deliberate corporate strategy rather than coincidence. The research is published as a working paper available via SSRN. Ma wrote that some companies in the study began cutting staff even before making AI investments, using the resulting savings to help fund those investments. Despite management’s expectation that both AI spending and job cuts would boost company value, Ma found that the average stock market reaction to layoff announcements was close to zero, and negative or flat for more than half of the events studied. Financial technology platform Block was cited as an exception, seeing its stock rise on news of AI-related staff cuts, according to the Wall Street Journal reporting referenced in the piece. To explain the disconnect, Ma’s team analyzed millions of employee reviews on Glassdoor.com and found AI-related comments were markedly more negative than the general tone of reviews on the platform. The research found a strong association between employee sentiment toward AI and firm productivity, and that sentiment dropped sharply following AI-related layoff announcements. Ma cited a Reuters/Ipsos poll finding that half of Americans fear AI could put someone in their household out of work. By contrast, an analysis of roughly 10,000 earnings-call transcripts found management’s tone on AI was consistently optimistic — but that optimism showed no significant relationship to actual productivity outcomes, according to Ma’s findings. The Real Story Isn’t AI Failing – It’s Management Using It as Cover Strip away the AI branding and this research describes something far more familiar: companies cutting costs and calling it innovation. The stock market’s near-zero reaction to AI-linked layoffs is the most telling data point in the entire piece, because investors, who have every financial incentive to reward genuine efficiency gains, mostly aren’t buying the story that these cuts are productivity-driven. If Wall Street isn’t rewarding it, and Ma’s own data shows productivity isn’t actually rising, the AI justification starts to look less like a technology story and more like a convenient narrative for decisions companies wanted to make anyway. What makes this genuinely dangerous, rather than just cynical, is the feedback loop Ma identifies: employees are told to master AI tools while watching their coworkers get laid off “because of AI,” and that fear measurably tanks the sentiment that actually predicts productivity gains. In other words, the layoffs aren’t just failing to produce the promised efficiency, they’re actively destroying the conditions AI would need to work at all. Any executive still leaning on “AI-driven restructuring” as a talking point in 2026 should have to explain why their own workforce’s trust, the one variable Ma’s research says matters most, is the first thing being sacrificed to get there.