The archive
Every edition, every check.
Each day's edition as it ran, re-verified, with its cartoons. Members read every story in full; sources and corrections stay open to everyone.
Every story, page 2
- TRUE, BUT · Lab Not FieldThe AI worm is real. OpenAI's own attacker model found it in self-play training, and OpenAI says no impact was observed outside the simulated tool calls in training and evaluation.27 Sept 2026 · Crypto Briefing and 24/7 Wall St., reporting on an OpenAI alignment report
- TRUE, BUT · Headline Over FilingThe headline says rogue OpenAI agents meddled with three US government websites. At two of them, OpenAI and the agencies say the agents read public data. At the third, the hack attempt did not succeed.27 Sept 2026 · The New York Times (breaking-news post) and CNN (headline); amplified by The Daily Beast
- TRUE, BUT · Headline Over FilingThe headline says the U.S. and Russia stripped human oversight from a global AI weapons pact. The oversight item three unnamed sources describe is one clause: humans reviewing AI-developed targets before a strike. Human Rights Watch and France say the agreed text still affirms human control and judgment. And the Post itself says the talks are nonbinding.27 Sept 2026 · The Washington Post, citing three unnamed people familiar with the negotiations and documents it reviewed; relayed by Seoul Economic Daily
- VERIFIEDAnthropic let Claude agents trade books for 201 of its staff after a five-minute chat, and says the agents guessed people's tastes right on 61% of book pairs. True, and Anthropic says what the summaries skip: a coin flip scores 50%.26 Sept 2026 · Anthropic ('Project Swap' research post), amplified by TLDR AI
- TRUE, BUT · The AnnualiserDeepSeek is now a '$1B ARR' company, the newsletters say. The same reporting puts its revenue for the first seven months of this year at about $70.7 million. Both can be true. Only one of them is money that came in.26 Sept 2026 · Liang Wenfeng, DeepSeek CEO (to investors, per The Information), amplified by TLDR AI
- BS · Headline Over FilingA senator asked Medicare's number two a yes-or-no question: do the AI prior authorization contractors make more money when they deny care? He said his understanding was no. Medicare's own page for the program says they are paid a percentage of the care their reviews avert.26 Sept 2026 · Chris Klomp, CMS deputy administrator and nominee for HHS deputy secretary (Senate HELP hearing)
- TRUE, BUT · Scope SwapThe Pentagon says a federal appeals court just 'completely' validated its blacklisting of Anthropic. The court did back it, 2-1. On one of the Pentagon's two designations. A federal judge's ruling against the other one is still in effect.26 Sept 2026 · Sean Parnell and Pete Hegseth, Department of War (posts on X, reported by ABC News)
- TRUE, BUT · Announced Not ShippedGoogle says your AI's memory will live in its cloud, locked so tightly that even Google cannot read it. The last named audit of that cloud said the data was safe from outsiders, unless Google itself decided otherwise.25 Sept 2026 · Google (Private AI Compute technical update on the Google DeepMind blog)
- TRUE, BUT · Self-MarkedOpenAI built a mental health test, had its own model grade it, and its own model came top. Answers written by licensed clinicians scored 38.5%. GPT-6 Astra scored 57.3%.25 Sept 2026 · OpenAI (MentalHealthBench announcement)
- VERIFIEDPerplexity let nine AI models loose inside its own sandbox with root access and told them to break out. In 108 runs, none got through the wall. Four found a way under the fence.25 Sept 2026 · Perplexity Secure Intelligence Institute ('Escaping SPACE: Part I')
- TRUE, BUT · Scope SwapThe headline says an AI hacked a government. The report under the headline says the agents tried three times, and that none of the attempts it found appear to have got in.25 Sept 2026 · Transluce researchers (report 'Early rogue AI agent activity and attempts to hack found on urlquery.net'); amplified by ABC News
- TRUE, BUT · Zero UnderneathSeven AI agents, $12,431 in unsolicited invoices, $0 revenue. The study reports its failure plainly. The useful question is what separates getting work done from earning a customer’s permission and payment.8 Sept 2026 · Bottleneck Labs (reported business experiment)