Anthropic CEO calls for China AI restrictions, not open model ban

Anthropic CEO Dario Amodei has clarified his company's controversial position on open-weight AI models, stating the firm does not support banning them while simultaneously calling for tighter restrictions on China's access to American AI technology.

In a statement published to Anthropic's website on July 27, Amodei addressed accusations that his company wants to eliminate open-weight models to protect its business interests. "We have not and are not advocating for a ban on open-weights models as a category," he wrote.

The clarification comes days after Anthropic notably declined to sign an industry letter opposing potential Trump administration restrictions on open-weight AI models. That letter, first reported by POLITICO, attracted signatures from over 130 companies including OpenAI, Nvidia, Microsoft, and Google.

Why it matters

The debate over open-weight models sits at the intersection of AI safety, national security, and economic competitiveness. Open-weight models from Chinese companies allow users to download and run sophisticated AI at significantly lower costs than proprietary offerings from Anthropic and OpenAI. Small businesses increasingly rely on these Chinese models, but U.S. officials worry about intellectual property theft and national security risks. Anthropic's isolated stance could influence whether the Trump administration pursues restrictions that would reshape the AI development landscape.

What Anthropic wants instead

Rather than model bans, Amodei outlined three policy priorities: blocking China's access to advanced semiconductors, combating Chinese intellectual property theft, and requiring mandatory safety testing for all "sufficiently capable" models.

The CEO directly challenged arguments made in the industry letter, which claimed that open-weight model access helps defenders develop safeguards against AI-enabled threats. "It seems at least as likely to me that the opposite will be true," Amodei wrote, suggesting that using AI to "weaponize pandemic-level viruses" could outpace defensive capabilities.

Industry backlash

Anthropicā€™s position has drawn sharp criticism from AI industry figures. Former White House AI and crypto czar David Sacks wrote on X that Anthropic's absence from the letter was "unlikely" to change, adding: "They won't stop until they kneecap open source. The rest of the industry needs to watch these guys like a hawk."

The Trump administration is currently weighing how to respond to increasingly powerful open-weight models and allegations that Chinese firms like Moonshot AI developed their products through large-scale attacks targeting American frontier models.

The industry letter argues that U.S. AI leadership depends on "whether the United States builds a strong, open ecosystem that diffuses into every sector," not just frontier model development.

These details were first reported by POLITICO.