Category Anthropic

Developments in AI safety and latest research from Anthropic.

Asian AI Startups Push Local Frontier Models as U.S. Export Curbs Open a Gap

AI export controls are pushing Asian startups to launch local frontier models as Sakana AI and 360 fill the gap left by Anthropic restrictions.

Margaret Atwood says AI reliability remains shaky after Claude gave her a wrong answer about Father Brown.

Claude growth is accelerating among paid consumers, with new data showing rising spend and interest even as ChatGPT still leads.

OpenAI’s first custom chip, Jalapeño, is built for ChatGPT inference and signals a broader push to reduce reliance on Nvidia.

Anthropic AI talks with the Trump White House are improving as Dario Amodei steps back and officials weigh whether to restore model access.

Figma AI tools now include motion, shaders and code layers, pushing its design canvas closer to a full-stack product workflow.

China and US AI experts are warning that frontier AI safety risks may require cooperation, even as geopolitical tensions rise.

AI researchers are leaving Google for Anthropic and OpenAI, intensifying concerns about talent retention in the frontier AI race.

Fitbit Air blends simple tracking with an optional AI health coach, showing how Google is trying to make Fitbit more useful without forcing AI.

Claude Tag brings persistent AI to Slack, giving Anthropic’s enterprise assistant shared context, memory and task support in channels.