Anthropic CEO Clarifies Company Stance Against Open-Weight Model Ban
Dario Amodei, CEO of Anthropic, clarified that his company has never advocated for a ban on open-weight AI models in response to growing industry criticism. This comes after a coalition including tech giants like Nvidia, Microsoft, Meta, and Palantir urged policymakers against premature restrictions on such models. OpenAI later joined the letter, but Anthropic did not sign it. Amodei emphasized that instead of banning open-weights models, efforts should focus on preventing authoritarian access to powerful AI chips, stopping industrial-scale distillation attacks, and ensuring safety testing for all advanced models—both open and closed. He also agreed with parts of the coalition's letter regarding customer control and competition but disagreed with claims about enhanced cybersecurity benefits from open-weight models. Amodei stressed that protectionist bans would not address his primary national security concerns, particularly in light of recent allegations against Chinese AI startups like Alibaba for distillation attacks on Anthropic’s Claude models.
Tickers: $META $MSFT $NVDA $PLTR
Topics: Artificial Intelligence Open-Weights Models Regulation Tech Company Statements