Keep pulling the thread on Mark Warner.
Senator Mark Warner reported that the National Security Agency (NSA) informed him that Anthropic's Mythos model was able to breach their classified systems during a red teaming exercise.
The U.S. government is delaying the public release of OpenAI's GPT-5.6 model, establishing an ad hoc licensing regime.
According to a memo from CEO Sam Altman, the U.S. government will approve access to OpenAI's GPT-5.6 on a customer-by-customer basis during its limited preview period.
OpenAI CEO Sam Altman announced internally that GPT-5.6 will initially be launched in a limited preview for a select group of partners instead of a broad public release.
The U.S. government has lifted its block on Anthropic's Mythos model for approximately 100 selected partners, including major U.S. companies and government agencies.
OpenAI has launched a new model named SOL, which is priced the same as GPT-5.5 and is described as a significant step forward in intelligence and efficiency.
OpenAI has launched a new model named TERRA, part of the GPT-5.6 family, which offers performance comparable to GPT-5.5 at half the price.
Micron reported strong earnings and projected that the structural supply chain shortage for AI components will continue for the foreseeable future.
Anthropic has released Claude Tag, a native integration of its Claude model into the Slack platform that utilizes a full instance of Claude Code.
Andrew Curran argues that the U.S. government's policy of delaying AI model releases widens the capability gap between internally held models and those available to the public, rather than slowing development.
Prediction markets at one point indicated a probability above 60% that Anthropic's "Fable 5" model would be re-released by July 1st.
The White House reportedly had more productive interactions with Anthropic co-founder Tom Brown after CEO Dario Amodei was sidelined from conversations regarding the company's AI models.