LATEST NEWS
Beacon Papers No. 12: The AI Race Needs a Trust Order
Beyond the False Choice Between Safety and Speed Two powerful arguments are shaping the future of AI. OpenAI and Anthropic warn that rapidly advancing capabilities demand stronger safeguards, independent evaluation, and coordination before humanity loses effective...
The One Thing AI Cannot Give Itself
Claude now leads a quarter of Anthropic's model R&D. Self-improvement is not self-authorization. Anthropic disclosed on September 17 that, as of August, about 90 percent of its model research and development tasks involved collaboration with Claude — the model...
