Beijing Rejects Anthropic CEO's Push to Curb China's AI Development
China pushes back against Anthropic CEO Dario Amodei's call to curb its AI capabilities, as Germany warns that halting development is unworkable and urges US-China involvement.
BEIJING / WASHINGTON — China’s government pushed back Sunday against a call by Anthropic chief executive Dario Amodei to restrict the country’s artificial intelligence development, escalating a public dispute over who sets the rules for advanced AI at a moment when no international framework exists to referee it, the Associated Press reported.
The exchange drew a counterpoint from Berlin the same day: Germany said that halting AI development altogether is not a viable path forward and urged the United States and China to participate in any governance conversations, Reuters reported.
Together, the statements expose a widening gap between American technology executives who argue that containing China’s AI capabilities is a legitimate national-security objective and a broader coalition of governments — including close U.S. allies — that view that framing as both unworkable and counterproductive.
Amodei’s Position
Amodei has argued in public settings that the United States and its allies should act to slow or constrain China’s access to the advanced AI infrastructure and expertise that could narrow or eliminate a Western advantage in frontier models. The argument mirrors the logic behind existing U.S. export controls on advanced semiconductors, which have sought to deny Chinese firms access to high-end chips used to train and run large AI systems.
For Amodei, the concern is partly commercial — Anthropic competes directly with Chinese AI developers for model dominance, talent, and deployment contracts — and partly strategic: advanced AI with military applications, from autonomous targeting to signals intelligence, could alter the balance of power between Washington and Beijing in ways that semiconductor controls alone cannot address.
Beijing’s Rejection
China’s government characterized the call as an effort by a U.S. technology company to use geopolitical levers to preserve a commercial and strategic advantage, the Associated Press reported. Beijing’s response follows a pattern in which it frames American AI restrictions — whether export controls, investment screening, or public advocacy campaigns — as economic protectionism dressed in security language.
The pushback carries diplomatic weight beyond the bilateral relationship. Several governments that have aligned with the United States on broader China policy have nonetheless resisted joining what they see as a technology blockade. Beijing’s public challenge to Amodei is partly aimed at those audiences: European governments weighing their own AI industrial policies, and Global South nations watching to see which superpower’s technology governance model they are being asked to adopt.
Germany’s Counterargument
Germany’s position, reported by Reuters, adds European institutional weight to the case against a restrictive approach. Berlin said that stopping AI development was simply not feasible and called on both Washington and Beijing to engage in multilateral governance talks rather than treating AI as a zero-sum competition to be won through denial.
The German stance reflects a broader European preference that has taken shape over several years: rules-based frameworks that include major AI powers rather than technology embargoes that exclude them. European governments have watched U.S.-China chip controls push Chinese firms toward domestic alternatives, a pattern they argue accelerates decoupling without achieving the security goals that motivated the controls in the first place.
Germany’s call for U.S. and Chinese involvement also signals frustration with a dynamic in which the two largest AI powers are effectively setting the terms of development — and the terms of the governance debate — without sustained input from a European bloc that has passed its own AI regulation and sees itself as a legitimate third voice in these conversations.
No Agreed Framework
The immediate context for Sunday’s statements is the absence of any durable international architecture governing AI development. Unlike nuclear weapons, chemical agents, or biological research with dual-use potential, frontier AI models face no treaty regime, no agreed verification mechanism, and no negotiated definition of what level of capability constitutes a threshold requiring international oversight.
Proposals for a global AI safety body have circulated in diplomatic channels, but none has produced binding commitments from the governments whose AI programs are the most consequential. The United States and China have held bilateral AI safety talks, but those conversations have operated at a level of generality that leaves the core disagreements — on access, on export controls, on military applications — unresolved.
Amodei’s call and Beijing’s rejection of it represent two poles of a debate that is proceeding in the absence of the institutions that would normally manage it.
Stakes for the Broader Rivalry
The dispute over AI development curbs sits inside a larger strategic competition that has intensified across multiple domains. China’s government recently amended national law to require that all domestic resources — economic, industrial, and logistical — be made available for military mobilization in a conflict, a change Defense News reported takes effect in October. AI systems, which have become central to modern military planning, logistics, and targeting, would fall squarely within the scope of that mobilization framework.
That context is part of what animates the U.S. argument for restrictions: that Chinese AI development is not a purely commercial enterprise separable from state military planning. Beijing’s position — that AI development is a legitimate national economic priority that foreign companies and governments have no standing to constrain — reflects an equally coherent strategic logic.
Neither position leaves obvious room for the multilateral framework Germany is calling for. But the alternative — two major AI powers developing increasingly capable systems under competing regulatory philosophies with no shared safety standards — is the status quo that Sunday’s exchange made more visible.
Related coverage: China Grows Wary as Putin-Kim Alliance Deepens | Trump, Iran, Oil: What the Venezuela Deal Signals | Kremlin Watches as Trump Proposes Ukraine Diesel Strike Deal
Found this useful? Share it.
