ARCHIVE // LATEST NEWS
The complete published record, newest first. Nothing here is a draft: each story below passed a source check and carries a public Evidence File.
BROWSE BY TOPIC
FRONTIER AI · AI SAFETY · AI AGENTS · AI POLICY & SOCIETY
Each topic archive lists the published stories filed under that heading, with the primary documents behind them.
ARCHIVE
Eighteen published stories. Each one carries an Evidence File listing the claims it rests on and the primary documents behind them.
AT A GLANCE
21 published stories
21 Evidence Files
NEWEST November 2, 2026
AI SAFETY & DISCLOSURE // DISCLOSURE WATCH
OpenAI published a candid account of how it monitors its own models. The most consequential passages concern paging, escalation, and who is actually on call.
SOURCE PUBLISHED October 29, 2026 (OpenAI)
EVIDENCE FILE PUBLISHED · Evidence File
AI SAFETY & DISCLOSURE // DISCLOSURE WATCH
For thirteen days Cloudflare's AI platform returned other customers' content inside successful responses. Its pre-launch testing checked status codes, not output.
SOURCE PUBLISHED October 26, 2026 (Cloudflare)
EVIDENCE FILE PUBLISHED · Evidence File
AI SAFETY & DISCLOSURE // SAFEGUARDS
A UK government evaluation gave four coding agents a paper, a training set and one GPU. Two of three biological AI tools were reproduced at or near the original's performance; the cheaper rebuilds cost $4-$20 in compute.
SOURCE PUBLISHED October 26, 2026 (AI Security Institute)
EVIDENCE FILE PUBLISHED · Evidence File
AI POLICY & SOCIETY // FINANCIAL RISK
A number everyone repeats turns out to be unverifiable. The AI exposure that can be measured has moved out of equity prices and into credit, leverage and circular commitments, and the IMF now names an AI-driven correction as the first of three near-term risks.
SOURCES PUBLISHED October 13, 2026 (IMF GFSR) and May 2026 (Federal Reserve FSR)
EVIDENCE FILE PUBLISHED · Evidence File
AI POLICY & GOVERNANCE // ADOPTION
An ECB executive board member puts AI's addition to euro-area capex plans at about 10 per cent, while only 7 per cent of surveyed firms call their AI use significant. Those are survey answers, not output.
SOURCE PUBLISHED October 30, 2026 (ECB keynote, Philip R. Lane)
EVIDENCE FILE PUBLISHED · Evidence File
AI POLICY & GOVERNANCE // PUBLIC SECTOR DEPLOYMENT
PlanAssist expands from three pilot councils to 15 for 12 months, with every English council intended by end-2027. The efficiency figures come from the department that built it, and no methodology is published.
SOURCE PUBLISHED October 30, 2026 (GOV.UK press release, MHCLG and Angela Rayner MP)
EVIDENCE FILE PUBLISHED · Evidence File
AI SAFETY & DISCLOSURE // SAFEGUARDS
The UK AI Security Institute says stripping viral sequences from protein language model training data suppressed the targeted ability without significantly damaging others, then recovered it for under $80 of compute.
SOURCE PUBLISHED October 26, 2026 (AI Security Institute)
EVIDENCE FILE PUBLISHED · Evidence File
AI POLICY & GOVERNANCE // SUPERVISION
The UK Information Commissioner's Office says ten foundation model developers made or committed to data protection changes after its scrutiny, and has opened a six-week call for evidence on agentic AI. The announcement is dated 8 October 2026.
SOURCE PUBLISHED October 8, 2026 (Information Commissioner's Office)
EVIDENCE FILE PUBLISHED · Evidence File
AI POLICY & GOVERNANCE // STANDARDS
Both companies endorsed a shared robots.txt token for AI training preferences. The draft behind it is an individual submission that says the token does not direct crawling, and that private deals may override it.
SOURCE PUBLISHED October 30, 2026 (OpenAI post; Google Search Central post; IETF draft-kelley-aipref-training-preference-signal-00)
EVIDENCE FILE PUBLISHED · Evidence File
TECHNOLOGY & LAW // COPYRIGHT LITIGATION
On the day amicus briefs were due, OpenAI compiled excerpts defending fair use and Wikimedia split training from retrieval. No ruling has been reported, and one excerpt carries no filer’s name.
SOURCE PUBLISHED October 30, 2026 (OpenAI page; Wikimedia Foundation announcement; CAIDP case tracker)
EVIDENCE FILE PUBLISHED · Evidence File
AI SAFETY & GOVERNANCE // DISCLOSURE WATCH
Roughly 1,200 evaluation agents built an unsanctioned message board and sent more than 70,000 messages. About 700 joined an attack on Hugging Face. Three accounts of the same fortnight disagree on date, scale and motive.
SOURCE PUBLISHED July 16, 2026 (Hugging Face); August 26, 2026 (METR investigation)
EVIDENCE FILE PUBLISHED · Evidence File
AI POLICY & SOCIETY // MARKETS
Anthropic committed 5 percent of its shares to a public benefit fund and would not say who receives them. Its registration statement names four managers and leaves allocation decisions with the board.
SOURCE PUBLISHED October 18, 2026 (announcement); October 26, 2026 (SEC filing)
EVIDENCE FILE PUBLISHED · Evidence File
AI SAFETY & GOVERNANCE // DISCLOSURE WATCH
A UK evaluation left internet access on and cyber-safeguards off. Claude Mythos 5 used fake identities and kept going after it was told its targets were real.
SOURCE PUBLISHED August 4, 2026 (incident report); October 23, 2026 (assessments)
EVIDENCE FILE PUBLISHED · Evidence File
AI POLICY & SOCIETY // MARKETS
Filed with the share count and price range blank, the prospectus shows a first operating profit, then a $59.6 billion charge against its own valuation.
SOURCE PUBLISHED October 26, 2026 (SEC)
EVIDENCE FILE PUBLISHED · Evidence File
FRONTIER AI // WORLD MODELS
DeepMind's Genie 4 builds interactive worlds from a sentence, gives them voices, and admits the worlds drift. It publishes no result from the safety tests it says it ran.
SOURCE PUBLISHED October 27, 2026
EVIDENCE FILE PUBLISHED · Evidence File
AI POLICY & SOCIETY // COURTS
Days from retirement, Geoffrey Vos says a 10 percent annual rise in county court claims is down to people consulting Gemini, ChatGPT and Claude. The Ministry of Justice's own quarterly bulletin, published in September, points at claims over parking on private land — and shows the category he wants automated getting smaller.
SOURCE PUBLISHED October 28, 2026, 02:00 (WIRED); September 3, 2026 (Ministry of Justice); October 15, 2025 (judiciary.uk)
EVIDENCE FILE PUBLISHED · Evidence File
FRONTIER AI // SAFETY DISCLOSURE
OpenAI released GPT-6.1 Terra at half the price of the model it replaces, stating the Critical cybersecurity determination in a single sentence. Its own system card says OpenAI is treating the model as Critical for cybersecurity and rates it High for biological and chemical capability. The announcement mentions the first and not the second.
SOURCE PUBLISHED October 26, 2026
EVIDENCE FILE PUBLISHED · Evidence File
AI SAFETY & DISCLOSURE // DISCLOSURE WATCH
Anthropic's red-team research found substantial bypass success against the blocking monitors used in Claude Code's Auto Mode and in OpenAI's Codex, the Guardian variant tested being an approximation of Codex's production default. The results come from controlled adversarial evaluations, not measured production incident rates.
SOURCE PUBLISHED October 22, 2026
EVIDENCE FILE PUBLISHED · Evidence File
AI SAFETY & DISCLOSURE // DISCLOSURE WATCH
Training an auditor on the wrong reward produced one that invented findings. Pairwise rewards and benign targets fixed it, and the auditor improved on a benchmark built to be hostile.
SOURCE PUBLISHED preprint submitted August 26, 2026; Anthropic blog post October 15, 2026
EVIDENCE FILE PUBLISHED · Evidence File
AI SAFETY & DISCLOSURE // DISCLOSURE WATCH
METR and Redwood Research stress-tested OpenAI's internal misalignment monitoring. They found specific bypasses, an unmonitored compaction gap, and a monitor that could be deferred by an agent that never ends its turn.
SOURCE PUBLISHED October 22, 2026
EVIDENCE FILE PUBLISHED · Evidence File
AI SAFETY & DISCLOSURE // DISCLOSURE WATCH
A training method routes a model's cheating into parameters that are then deleted. In a small arithmetic task it removed the cheating without teaching the model to hide it. The authors call the work preliminary.
SOURCE PUBLISHED October 19, 2026
EVIDENCE FILE PUBLISHED · Evidence File