Research AI Minute Newsroom 2026-08-24

Roughly 90 percent of executives say AI has not raised productivity. The cuts are happening anyway.

Mark Ma, a professor of business administration at the University of Pittsburgh, published an analysis on 12 August, picked up widely on 22 August, built on five years of US public-company data: millions of employee reviews from Glassdoor, thousands of company financial reports, hundreds of AI investment and layoff announcements, and roughly 10,000 earnings-call transcripts. He anchors it on a separate finding from an Atlanta Federal Reserve survey — about 90 percent of executives say AI has not yet raised productivity at their company. His own results are the uncomfortable part. The more often a company announces AI investment, the more often it announces job cuts it attributes to AI. The stock market barely responds: the average return around an AI-linked layoff announcement is close to zero. Inside the companies, AI-related comments in employee reviews are markedly more negative than the overall tone of those reviews, driven by job insecurity, a lack of training and poor leadership on AI. And the association that does hold is between positive employee sentiment toward AI and higher firm productivity. Management optimism showed no such relationship.

Why it mattersThe AI-layoff story is usually told as a hard-headed efficiency trade: cut people, keep output, bank the difference. This data says the trade is not showing up in the numbers — not in productivity, where nine in ten executives report no gain, and not in the share price, where investors are treating the announcements as roughly information-free. If you are an employee being reorganised around these tools, the finding to take is the one about sentiment: the measurable link runs from how workers feel about AI to how productive the firm is, not from how convinced the boss is. That points at a boring, unglamorous answer — training, and some assurance about jobs — which is exactly the spending that gets cut when a company is funding AI by cutting.

✓ Verified · 2 sources

Read in the app — free, in 9 languages

An older Claude broke Anthropic's own content rule in 10 tries out of 10 — and it is still sold through Azure and Bedrock
2026-08-23 When a small model hands a conversation to a big one, the big one re-reads everything. Nvidia says a straight line can do the translation instead.
2026-08-23 A 27-billion-parameter model beat Opus 4.8 and GPT-5.5 at reproducing published papers
2026-08-23 Give a model the reference list of an unpublished paper and ask what the paper says. The best ones get it 15 percent of the time.
2026-08-22 Three-quarters of Americans do not want a data centre near them. A year ago they were evenly split.
2026-08-22

Roughly 90 percent of executives say AI has not raised productivity. The cuts are happening anyway.

Related stories