Anthropic’s founder and CEO, Dario Amodei, addressed industry speculation on Monday afternoon regarding claims that his company backs U.S. government efforts to ban open-weight Chinese AI models—or potentially all open-weight models. “Anyone who has read my past writing should know that I don’t regard such bans as a useful measure, but let me state it clearly so that there is no doubt: Anthropic has never advocated for a ban on open-weights models,” he wrote, emphasizing his own words. This clarification came after Nvidia’s founder and CEO, Jensen Huang, made his first post on X on Friday, sharing an open letter. In that letter, Nvidia, along with a broad coalition of AI companies—including Hugging Face, Meta, Microsoft, Mistral, and Nvidia itself—urged policymakers to avoid imposing sweeping “premature restrictions” on open-weight AI models.

Although the letter didn’t specifically reference China, the ongoing industry debate has revolved around allegations that Chinese AI labs are advancing rapidly, often by stealing intellectual property from U.S. counterparts. One method cited is distillation, where an AI system bombards another model with queries to learn its inner workings. In a blog post published Monday, Amodei argued that open-weight models fall into a different category than the threat posed by China. “Open-weights models that don’t have dangerous capabilities are a public good: they don’t cost anything besides the compute needed to run them, and they provide value to businesses, developers, and researchers,” he stated.

Amodei noted in his post that he has long-standing concerns about AI, but the use of open-weight models by businesses—even those from China—is not among them. His real fear is that “authoritarian governments” could build models more powerful than those in the U.S., aiming for “permanent military superiority” or to repress their own populations. While he acknowledged that the Chinese Communist Party (CCP) isn’t the only authoritarian government he worries about, he described it as the “most capable.”

EMBED_PLACEHOLDER_0

Amodei also expressed fears that AI could enable “biological attacks,” not just cybersecurity threats. In his view, open-weight models are particularly dangerous in these scenarios because it’s difficult to apply guardrails or monitor their usage. He referenced a UK AI Security Institute report, noting that once open-weight models are released, they cannot be recalled. This perspective contradicts what open-source advocates argue: that access to powerful open models, not controlled by a single entity, helps defenders protect themselves.

Amodei outlined actions he believes would counter China, such as restricting its access to powerful chips—a long-standing U.S. policy—and calling for a formal crackdown on distillation. The U.S. has already threatened sanctions against China if it finds evidence of IP theft involving U.S. models. Interestingly, Amodei also voiced support for growing efforts, some led by the U.S., to establish a model safety testing organization, especially if the entire world—including China—agrees to participate. “I think this idea is actually close to a consensus: I have been heartened both that the Trump administration has moved in this direction in recent months,” Amodei wrote. “Note that to be effective, testing would need to be global, which means even the CCP would need to be on board. I think this may actually be possible … limited cooperation around preventing AI biological weapons may be possible because it is in China’s interest too.”