Tag: Anthropic
Anthropic traces a 200M-exchange campaign to copy Claude’s thinking
A new Anthropic threat report describes five distillation operations that harvested Claude's reasoning traces, led by a 151M-exchange…
A botched sandbox test left an AI agent fighting a login puzzle
Anthropic's September misuse reporting pairs a sandbox escape with a 1,022-page transcript of a model losing to an…
Anthropic says infostealer gangs are draining Claude accounts
Anthropic tells users that infostealer malware is behind a wave of drained Claude accounts.
Jacob Coxon leaves Anthropic over self-improving AI fears
A researcher who trained models at OpenAI and Anthropic has quit, calling the race to self-improving AI a…
Navier-Stokes proof sparks a fight between OpenAI and two rivals
NYU and Anthropic researchers accuse OpenAI of jumping their math work after hearing about it.
Anthropic’s UK hire steps back from ARIA leadership within days
Britain's ARIA founding chair is stepping aside five days after taking a government affairs job at Anthropic.
Anthropic open-sources a blueprint for AI commerce agents
Anthropic published an Apache-2.0 reference design for shopping and merchant agents built around a single skilled agent.
Anthropic paused risky training runs for weeks after agent incidents
The lab suspended external cyber evaluations and high-risk reinforcement learning after unauthorized actions by its agents.
Pentagon insists Anthropic stays a supply chain risk after court defeat
Under Secretary Emil Michael reaffirmed the blacklisting on X days after a judge voided it and Lutnick said…
ChatGPT, Claude and Grok go down together in a rare triple outage
All three leading AI assistants suffered degraded service at once on September 3, with OpenAI logging 19 broken…
Anthropic ships Claude Fable 5.1 wide while Mythos 5.1 stays gated
The same core model splits into a general release and a vetted cyber and science edition.
China’s state TV outlet sets conditions for US AI safety talks
Chinese state media sets conditions for US-China AI safety talks in an attack on Anthropic.