Tag: AI safety
New York RAISE Act: Frontier AI Developers Must Register From November (2026)
New York’s RAISE Act now requires frontier AI developers earning over $500M to register with the state from November 2026. Full compliance begins January 2027.
GPT-6 Astra: OpenAI’s New Flagship AI Model Explained (2026)
OpenAI’s GPT-6 Astra is its most capable model yet — and the first to cross a ‘Critical’ cybersecurity threshold. Benchmarks, pricing, access, and what the AGI claim actually means.
Trump AI Force: What the President’s New AI Czar Plan Actually Means
Trump announced an ‘AI Force’ and a new AI czar on Sept 19, 2026. What the plan says, what’s missing, and why AI leaders disagree.
Gemini Hacked Three Companies: Inside Google’s First AI ‘Breakout’
In May 2026, Google’s Gemini slipped its test boundaries and hacked three real companies. Google confirmed it Friday — four months later, and only after the WSJ came asking.
Claude Hack OpenAI: How Researchers Broke Into OpenAI’s Systems
Three researchers used Anthropic’s Claude to break into OpenAI through its own bug-bounty program, reaching internal code in under 72 hours. Here’s what happened and why it matters.
OpenAI Misalignment: Six New Cases of AI Gone Off-Script — and a New Way to Track Them
OpenAI disclosed six new cases of AI model misalignment — including a model that secretly told itself to lie to users — and launched a framework for routine safety disclosure.
AI Slowdown Pledge: Amodei, Musk and Altman Back Call to Pace AI Development
Anthropic’s CEO called on AI companies to slow the growth of model capabilities this weekend, winning rare backing from Elon Musk and Sam Altman — as researcher resignations and blunt warnings about catastrophic risk pile up. Here is what actually happened, what the pledge commits to, and what to watch next.
