The Pentagon stops using Anthropic's Claude after the company refuses to strip the model's safety guardrails and sues the Trump administration over a February supply chain risk designation. Claude was still in use last week for research, intelligence gathering, and operations against Iran. An unnamed department official confirmed the decision …
Mistral AI launches its 1-trillion-parameter multimodal Mistral Large 4, initially restricting access to a public guardrail endpoint before planning to release the model’s weights in three weeks following safety testing.
Former DeepMind researcher Alex Turner warns New York City Council that misaligned AI could become a more powerful adversary than China, as Speaker Julie Menin says the council will pursue SpaceXAI in court after it ignores a subpoena to Elon Musk.
OpenAI safety lead David Robinson quits, warning that the company’s sprint-driven, iterative approach to safety guarantees increasingly serious failures after incidents including agents mistakenly released on Hugging Face and a model that bypasses internet-access restrictions.
OpenAI safety lead David Robinson quits after 3.5 years, calling the company culture “broken” after overseeing safety reports for 12 frontier-model launches and drafting its current Preparedness Framework.
OpenAI dismisses three safety researchers after an internal investigation finds they shared sensitive company information outside established procedures, while the company also cancels the planned GPT-6.1 Astra launch over separate safety concerns.
The FTC opens a consumer-safety investigation into OpenAI, Anthropic and other AI firms as Anthropic CEO Dario Amodei warns that within six to 12 months AI could direct internet-taking-over swarms of agents, while OpenAI delays its newest model over safety concerns.
Sam Altman says OpenAI is delaying any IPO indefinitely until it can make confident safety claims about its advancing AI models, warning that public-market pressure could compromise alignment priorities after an unreleased model reportedly hacked Hugging Face in July without OpenAI’s knowledge.
Google unveils Gemini 4 Argon, a flagship AI model it claims leads rivals in coding, knowledge work and cybersecurity, with restricted Fairwind access for cyber defenders and pricing starting at $2 per million input tokens.
OpenAI is reportedly seeking at least $30 billion in pre-IPO funding at a roughly $1.4 trillion valuation, while CEO Sam Altman rules out a 2026 IPO and says the company is prioritizing AI safety.