Anthropic CEO Criticizes US Approval for Nvidia AI Chip Exports to China
Anthropic CEO Dario Amodei strongly criticized the Trump administration's approval of Nvidia AI chip exports to China during a Davos panel, comparing it to selling nuclear weapons to North Korea. The new policy allows sales of Nvidia's H200 chips with third-party review and quantity limits, reversing Biden-era restrictions. While former officials warn of boosting China's AI capabilities, the administration argues controlled exports could preserve US market dominance and discourage Chinese competitors.

*this image is generated using AI for illustrative purposes only.
Anthropic CEO Dario Amodei has issued a stark warning against the Trump administration's recent decision to allow Nvidia to sell advanced AI chips to China, drawing a provocative comparison to selling nuclear weapons technology to North Korea. The criticism came during a panel discussion on artificial general intelligence at Davos, where Amodei expressed serious concerns about the national security implications of the policy shift.
Strong Opposition from AI Industry Leader
Speaking at the World Economic Forum, Amodei used pointed language to illustrate his concerns about the export decision. "Are we going to, you know, sell nuclear weapons to North Korea, and, you know, because that produces some profit for Boeing?" he said, according to AP reports. The Anthropic chief executive warned that shipping such powerful chips risks providing China with the computational power necessary to develop super-intelligent AI systems.
| Concern Area: | Details |
|---|---|
| Risk Assessment: | Handing China compute power for super-intelligent AI |
| Security Impact: | Potentially far-reaching national security implications |
| Industry Comparison: | Nuclear weapons technology transfer |
Broader Industry and Policy Concerns
Amodei is not alone in his criticism of the new export policy. Saif Khan, who previously served as director of technology and national security on the White House National Security Council under Joe Biden, has also warned that the rule would substantially boost China's AI capabilities. The concerns reflect broader anxiety within the AI community about the geopolitical implications of advanced chip technology transfers.
The Anthropic CEO's comments came during discussions about artificial general intelligence, systems that reportedly have the capability to outperform human intelligence. Amodei noted that while AI companies are racing to develop AGI, such technology also carries risks that could threaten humanity.
Trump Administration's New Export Framework
The Trump administration approved China-bound sales of Nvidia's second-most powerful AI chips last week, implementing a new regulatory framework that will likely enable shipments of the H200 model. This decision represents a significant shift from Biden-era restrictions that had barred sales of advanced AI chips to China.
| Policy Details: | Specifications |
|---|---|
| Chip Model: | Nvidia H200 (second-most powerful) |
| Review Process: | Third-party testing lab verification |
| Quantity Limit: | Maximum 50% of chips sold to US customers |
| Previous Status: | Banned under Biden administration |
Under the new regulations, chips will be reviewed by a third-party testing lab to confirm their technical AI capabilities before export approval. Additionally, China cannot receive more than 50% of the total amount of chips sold to American customers, creating a built-in limitation on the volume of exports.
Administration's Strategic Rationale
Despite the criticism, the Trump administration maintains that the policy shift serves strategic purposes. White House AI czar David Sacks believes that allowing controlled exports of Nvidia and AMD chips could discourage Chinese competitors, particularly heavily sanctioned companies like Huawei, from accelerating efforts to develop competing advanced chip designs. The administration argues this approach could preserve US companies' market dominance while managing competitive threats.
The debate highlights the complex balance between commercial interests, technological leadership, and national security considerations in the rapidly evolving AI landscape. As companies like Anthropic continue developing increasingly powerful AI systems, the question of technology transfer and export controls remains a critical policy challenge.
























