Anthropic CEO denies support for ban on open-weight AI models

Anthropic chief executive Dario Amodei said his firm has never pushed for a ban on open-weight artificial-intelligence models, responding to criticism that it supported a letter signed by Nvidia, Palantir and other AI firms. In a post on Anthropic’s website, Amodei wrote that he does not view such bans as useful and emphasized that the company sees non-dangerous open-weight models as a public good that benefits businesses, developers and researchers.

He acknowledged agreement with much of the open letter’s emphasis on expanding access to the AI economy and strengthening competition, but disagreed with the claim that open-weight models inherently make it easier to develop safeguards. Amodei argued that the focus should instead be on preventing powerful chips from reaching authoritarian regimes, stopping large-scale model distillation, and requiring safety testing for all capable models, whether open or closed.

The response was issued after the Open Secure AI Alliance, a coalition that includes Nvidia and Palantir, announced a partnership aimed at improving AI security. Palantir chief executive Alex Karp, whose company signed the alliance’s letter, warned that restrictive AI regulations could hinder competition and urged the U.S. administration not to ban open models. > "If you want to win, you have to compete on the battlefield," Karp said in an interview.

Amodei’s statement seeks to clarify Anthropic’s stance amid ongoing debate over how to balance openness and safety in the rapidly evolving AI sector.