Anthropic answered the open-weights letter on Monday, three days after it published. "Anthropic has never advocated for a ban on open-weights models," CEO Dario Amodei wrote in a company post.
Anthropic answers the open-weights letter: 'never advocated for a ban'
Dario Amodei agrees with much of the letter but rejects its claim that openness favors defenders. Anthropic still has not signed. The list hit 133 names Monday, and Amazon is now on it.
It still has not signed. The letter reached 133 signatories by Monday evening, and Amazon is now among them. Anthropic and Apple remain the notable absences.
Why it matters: The letter has become a public register of where each AI company stands on open models. Microsoft, Meta, NVIDIA, Google, OpenAI and now Amazon are all on it. Anthropic chose a counter-position over a signature.
What Amodei says: Open-weight models without dangerous capabilities "are a public good," he writes. He ties the letter to reports that US officials are considering banning American companies from using Chinese open-weight models, and notes some critics have accused Anthropic of "wanting to ban open-weights models as a means of protecting our business."
A ban on their use, he writes, "would protect US AI companies from competition, but that has never been my goal."
Instead of a ban, he backs three measures Anthropic has pushed before: blocking powerful chip sales to China, cracking down on "industrial-scale distillation operations," and mandatory safety testing for "all sufficiently capable models, open and closed."
Where he splits from the letter: "I agree with much of it," Amodei writes: open weights expand access to the AI economy and strengthen competition, at least for some use cases. What he rejects is the letter's assertion that open weights necessarily make it easier to develop safeguards or help defenders more than attackers. "It seems at least as likely to me that the opposite will be true," he writes, pointing to biology, where he worries the attacker holds the structural advantage.
