AI firms Anthropic and OpenAI make contrasting moves amid open‑source model debate

Anthropic, a U.S. artificial‑intelligence startup, declined to sign a petition urging support for open‑source AI models. The company defended its closed ecosystem as safer and accused Chinese firms of “mining” its Claude models for training. CEO Dario Amodei said, “Anthropic does not encourage the banning of open AI models… Open models without dangerous capabilities are public capital,” while critics argue the firm’s rhetoric is meant to sway regulation.

OpenAI announced the return of former researcher Lilian Weng, who left the startup Thinking Machines for health reasons. Weng will head a new research group tasked with accelerating OpenAI’s AI research. OpenAI research director Mark Chen welcomed her back, stating, “Weng will lead a research group to speed up OpenAI’s AI research.”