Silicon Valley’s AI Strategy Shifts From Competition With China to Calls for Cooperation

For years, some of Silicon Valley’s most powerful leaders have argued that the United States is engaged in an artificial intelligence arms race with China—and that losing could have dire consequences. Over the past week, however, those same executives have emphasized a different message: AI could be extremely dangerous, and the two countries urgently need to work together to slow the pace of competition.

Anthropic CEO Calls for a Slower AI Race

“To set the pace for the world, we need to work with China, an authoritarian nation with by far the most advanced AI capabilities,” Anthropic CEO Dario Amodei wrote in a blog post published Saturday, “We Must Pace the Frontier.”

The Chinese government and AI companies have made clear that they are deeply skeptical of Amodei’s proposals—not because they want to continue competing blindly or because they dismiss the technology’s serious risks, but because they believe they are being offered a bad deal.

“China’s AI safety community is growing and shares many of the same concerns as American labs,” said Kevin Hsu, founder of technology hedge fund Interconnected Capital and a former White House staffer under President Barack Obama. “There is a gap between a human-led vision of only conquering and limiting China first and maximizing its influence before engaging in dialogue. On the other hand, China will only engage in dialogue if it is treated as an equal.”

Why China Rejects the Proposed AI Cooperation Model

Amodei wrote that a “key part” of his slowdown strategy is for the United States and other democracies to maintain “as large a lead as possible” in AI development over authoritarian states such as China. He called on the United States to take measures including cracking down on chip smuggling and preventing Chinese companies from distilling unlicensed AI models.

“His paper specifically states that the pace of AI is a condition for widening the gap with China,” said Alvin Wang Graylin, a senior fellow at the Center for China Analysis at the Asian Social Policy Institute and a digital fellow at the Stanford Institute for Human-Centered AI. “Therefore, this proposal will feel inauthentic to Chinese readers.”

Amodei also argued that weakening China’s AI industry could increase U.S. influence and “increase the likelihood of a future agreement.” Exactly how that approach would bring the Chinese government to the negotiating table remains unclear. Amodei later acknowledged in a television interview that he had “no idea how to negotiate with China,” as he admitted.

China Calls the Proposal “Fear-Mongering”

Chinese government officials, AI researchers and other prominent commentators who responded to Amodei did not argue that artificial intelligence poses no risks. Many, however, clearly resent the idea that China should be treated both as an important governing partner and as a competitor to be contained.

When asked about Amodei’s blog post at a press conference on Monday, Chinese Foreign Ministry spokesperson Guo Jiakun offered a blunt response. He described the proposal as “fear-mongering, divisive and viciously competitive,” arguing that it would hinder global AI cooperation. The same day, the ministry emphasized that “China attaches equal importance to the development of AI and national security.”

AI Safety Concerns Differ in the United States and China

The mixed reaction reflects another reason China’s government may be less receptive to Anthropic’s warnings than other audiences. Unlike many of Silicon Valley’s loudest voices, Chinese leaders are not primarily focused on the so-called existential risks of AI, such as the possibility that rogue agents could kill everyone on the planet.

As in other policy areas, a pervasive concern within the Chinese government is that AI could disrupt domestic social stability or threaten the Chinese Communist Party’s grip on power.