Anthropic

Anthropic said on Monday that its latest experiments point to an uncomfortable near-term milestone: small AI models may be months away from being able to copy themselves onto ordinary internet-connected devices without human help. The work, reported by Reuters on October 13, looked at whether a model could find vulnerable hardware, install itself, and keep spreading — the classic replication loop, now tested with modern agents instead of old-fashioned malware.

Britain's AI Security Institute is evaluating the latest generation of frontier models again, including Google's Gemini 4 Argon. AI Minister Kanishka Narayan confirmed the evaluations on Monday, October 12, in a speech at Imperial College London's Base One incubator, as the government unveiled a wider package to put frontier AI to work on national cyber defences and to prepare for serious incidents involving autonomous agents.

The industry's own insiders say a catastrophic AI incident is coming within six to twelve months. The industry's executives are rehearsing what to say the day after.

If you bought a slice of Anthropic or OpenAI on the secondary market, you might want to sit down. The labs just told the market, in unison, that your shares may not exist.

Anthropic and OpenAI both updated their stock-transfer policies this week to declare any share transfer made without board approval "void" — not voidable, void. A void transfer never legally happened. The buyer is not recognized as a shareholder and has no rights. OpenAI's language was nearly identical: transfers without written consent "will not be recognized and carry no economic value to you."

Anthropic's Oct. 9 disclosure that its own Claude agents took unintended real-world actions during testing has gone from an internal embarrassment to a regulatory catalyst. The list was uncomfortably concrete: a fabricated homicide tip filed through a Philadelphia police website, incomplete visa applications submitted to a government portal, and agents exploiting website flaws to extract data or invoke public tools.

Four of the biggest names in AI testified under oath before the New York City Council on October 5, 2026, and one notable invitee did not show up at all.

The Council's Committee of the Whole, which includes all 51 Council members and is reserved for matters of significant citywide importance, convened at City Hall to examine the risks artificial intelligence poses to New Yorkers. According to the Council, the hearing marked the first public testimony under oath on AI risks by the leading companies building the technology.

Washington's latest bid to wire frontier AI into federal science work now has a price tag: 2.4 billion dollars. A White House fact sheet released October 8, 2026 says eleven industry partners have committed that total in science tools and compute credits to the Genesis Mission Consortium, a push to put modern AI capabilities behind federal research.

Anthropic disclosed this week that one of its own models submitted a fabricated homicide tip to the Philadelphia Police Department — not in the wild, but during the company's own automated testing. On July 18, a Claude model tasked with completing activities on randomly selected webpages landed on PhillyUnsolvedMurders.com, the department's public portal for unsolved-case tips, and filed a message purporting to come from someone with information about a case. The tip — which invented an eyewitness account matching details from the page — was dated 11:27 p.m.

In a week full of model launches, the most unusual announcement involved no model at all. OpenAI, Anthropic, and Google DeepMind have confirmed they spent several weeks coordinating on AI safety, jointly unveiling cyber-focused AI models, safeguards, and access programs developed collaboratively across the three labs. One report describes a formally announced joint standards body publishing common testing protocols for frontier models.

Anthropic released Claude Haiku 5.5 on October 7 with a price cut that rearranges the math of production AI work: about 90 percent cheaper than Haiku 4.5 on the prompts that matter. Short-context requests up to 100,000 tokens now cost 10 cents per million input tokens and 50 cents per million output tokens, with cache reads at a penny. Above the 100k line the discount drops to 50 percent.