Anthropic’s Dario Amodei Responds: Doesn’t Oppose Open-Weight Models, But Fears Chinese AI
Anthropic CEO clears the air on industry rumors, emphasizing that harmless open-weight models are a public good while outlining measures to counter China.
Stock photo for illustration only, not from the actual event
- Anthropic CEO Dario Amodei dismissed rumors that his company supports banning open-weight models.
- Jensen Huang led an open letter with major tech firms urging policymakers not to prematurely restrict open-weight AI.
- Amodei views safe open-weight models as a public good that benefits businesses and researchers.
- He proposed targeting advanced chip access and distillation crackdowns rather than banning open models.
Dario Amodei, the founder and CEO of Anthropic, responded on Monday afternoon to industry murmurs suggesting that his company somehow supports U.S. government efforts to ban open-weight Chinese models, or open-weight models generally.
His response followed an open letter shared on Friday by Nvidia founder and CEO Jensen Huang—marking his first post on X—alongside a long list of AI companies including Hugging Face, Meta, Microsoft, and Mistral, urging policymakers to avoid imposing broad premature restrictions on open-weight AI 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"
Dario Amodei
Although that letter did not specifically mention China, industry debates have centered on allegations that Chinese AI labs are growing their capabilities often by stealing intellectual property from American counterparts through distillation, a method where an AI bombards another model with prompts to learn its inner workings.
In a blog post published on Monday, Amodei stated that he views open-weight models as falling into a completely different category from the threat posed by China. He noted that open-weight models without dangerous capabilities serve as a public good, costing nothing beyond the compute needed to run them while providing immense value to businesses, developers, and researchers.
Stock photo for illustration only, not from the actual event
The debate surrounding open-weight AI models highlights a deep divide in the tech industry. Proponents of open source argue that broad access to powerful models allows defenders to secure their own systems, whereas national security advocates worry about technology proliferation. Amodei's stance attempts to decouple the open-source software debate from geopolitical competition involving state actors.
This viewpoint directly contrasts with arguments from open-source advocates who contend that access to powerful open models not controlled by a single entity helps defenders protect themselves. In response, Amodei listed specific actions he believes would effectively thwart China.
His proposed countermeasures include restricting China's access to powerful chips, which has been a longstanding U.S. policy, and calling for a formal crackdown on distillation. The U.S. has previously threatened sanctions against China if intellectual property theft involving American models is detected.
Furthermore, Amodei mentioned his support for growing efforts, some led by the U.S., to establish a model safety testing organization, especially if the entire global community, including China, agrees to submit to it.
Source: TechCrunch
Found something wrong in this article? Report an issue with this article
Comments
Leave a Comment