AI-safety

Overnight lobbying by Elon Musk, Mark Zuckerberg, and David Sacks persuaded President Trump to cancel a planned executive order that would have let federal agencies preview frontier AI models up to 90 days before release.

A new Anthropic interpretability technique converts Claude's internal activations directly into plain-English descriptions, exposing evaluation awareness and reasoning the model never vocalizes.

A coming White House executive order would require AI labs to share frontier models with the government 90 days before public release, but infighting between economic and national security factions has delayed the response to Anthropic's Mythos.

OpenAI launched GPT-5.5-Cyber on May 7, 2026 to vetted security teams via its Trusted Access for Cyber program, after CEO Sam Altman publicly criticized Anthropic's restricted Mythos rollout.

The Commerce Department's AI standards center will evaluate the three labs' frontier models in classified environments before public release, expanding a program that began with Anthropic and OpenAI in 2024.

Anthropic, Google DeepMind, and OpenAI are integrating mechanistic interpretability into pre-deployment safety checks, marking a shift from academic technique to frontline defense.

Governor Hochul's March 27 chapter amendment transforms New York's RAISE Act from a vague skeleton into binding law, imposing 72-hour incident reporting and mandatory safety frameworks on frontier AI developers before federal preemption efforts can intervene.

OpenAI releases a three-pillar framework developed with NCMEC, Thorn, and state attorneys general to combat the surge in AI-generated child sexual abuse material.

Executive Order N-5-26 requires AI vendors seeking California state contracts to certify safeguards against illegal content, bias, and civil rights violations within 120 days.

Nearly 3,000 unpublished assets exposed through an unsecured content management system reveal Anthropic's next-generation AI model and its own warnings about its offensive cyber capabilities.

The second International AI Safety Report, led by Yoshua Bengio with 100+ experts, warns AI capabilities are outpacing safety measures as frontier models show signs of deception.

OpenAI releases its most capable model yet with autonomous computer operation and enterprise integrations, but the launch is overshadowed by Pentagon deal fallout and a reported 1.5 million user exodus.