Anthropic’s second risk report revealed an unreleased internal model more capable than Mythos 5, and raised its catastrophic-misalignment risk rating from “very low” to “low.”
The US is preparing to warn dozens of countries they must choose between its Pax Silica AI coalition and China's competing framework, or risk exclusion, according...
OpenAI disbanded its Preparedness team, which assessed whether its models could enable catastrophic harm, distributing the work across other teams as it prepares for a major...
Investors expect Anthropic to target a valuation of $2 trillion or more in an October IPO, which would be the largest listing in history, driven by...
Alibaba released the open weights for Qwen3.8-Max, a 2.4-trillion-parameter model it says trails only Anthropic's Fable 5, its first openly released flagship-class system.
Suspected China-linked hackers used a team of open-source AI agents to run a near-autonomous cyberattack on Taiwan's government, breaching 85 accounts in what researchers call a...
Researchers found a flaw letting them extract the hidden reasoning of Claude, ChatGPT and Gemini by replaying encrypted traces into weaker sibling models, exposing secrets and...
Spotify will add "AI Persona" badges to AI-generated artist profiles from mid-September and exclude their music from recommendations, as AI tracks flood streaming services.
Cloudflare launched Kitesurf, a cloud browser designed for AI agents rather than humans, stripping out tabs and visuals to use up to seven times less memory...
An unreleased Claude model raised a longstanding lower bound tied to the Riemann hypothesis from 41.6% to 67.2%, a validated result that stops well short of...
Anthropic will embed an imperceptible watermark in text from new Claude models, a mark that survives copy-paste and some editing, to comply with EU AI transparency...
Anthropic has reportedly signed a 20-year, $9.1 billion data center agreement with Riot Platforms for 191 MW of AI computing capacity at the bitcoin miner’s Rockdale...
OpenAI is replacing GPT-5.5 Instant with GPT-5.6 Luna for Free and Go users, then removing text-chat limits and adding a Think button for harder questions.
