US and China Seek Common Ground on AI Safety Ahead of Trump-Xi Meeting

As concerns mount over the dangers posed by artificial intelligence, the prospect of any coordinated global approach rests heavily on cooperation between the United States and China — two powers whose mutual suspicion of each other's AI ambitions appears only to be deepening.

AI governance is expected to figure in a planned meeting between President Donald Trump and Chinese leader Xi Jinping in Washington next week. The issue has acquired new urgency in both capitals, with American technology leaders arguing that preventing worst-case scenarios involving rogue AI will require some level of engagement with China.

Yet both nations are simultaneously competing for a decisive edge in the technology. Trump has warned that regulating AI too aggressively could hand China an advantage, and has repeatedly said the country that leads in AI "wins." In a social media post on Monday, he asserted that the United States is ahead of China and all others.

The two countries are advancing along different paths. American firms are pushing the technological frontier with cutting-edge chips and a rapidly expanding data center network, while China is promoting broad global adoption of its open-source, lower-cost models. Chinese AI models have made significant strides and are gaining users among American companies, even as US-led restrictions block China's access to the world's most advanced AI chips. Treasury Secretary Scott Bessent said Tuesday that the US needs to develop more open-source models of its own to counter China.

Washington has accused Beijing of "aggressive, malicious" attempts to extract capabilities from American AI systems. China has responded by accusing the US of trying to suppress Chinese companies in order to build a "monopoly of the AI industry."

Whether the personal relationship between Trump and Xi is strong enough to bridge those differences remains uncertain. But figures with ties to the Chinese government see value in coordination. Speaking Tuesday at the Xiangshan Forum, an annual security and defense conference in Beijing, former Chinese ambassador to the US Cui Tiankai said AI is an area where dialogue between the two capitals "should go ahead and build up," citing a growing need for cooperation.

Two leading American AI companies, Anthropic and OpenAI, restrict access in China and have described protecting the US lead in AI as a national security imperative. At the same time, both have called for cooperation with China on safety management. In an essay last weekend urging a slowdown in AI development for safety reasons, Anthropic CEO Dario Amodei said the US should slow China's progress, including by maintaining restrictions on its purchases of advanced AI chips. He also wrote that "global pacing will require cooperation with China, the autocratic country with by far the most advanced AI capabilities."

The essay drew swift criticism from Beijing. China's Ministry of Foreign Affairs said Monday that "fearmongering, confrontation and vicious competition will only disrupt the process of global AI governance which serves no one's interest." The state-run Global Times went further, describing his remarks as "packed with containment provisions targeting China" and, in essence, a "Cold War playbook" for the AI sector.

Trump has consulted with Nvidia CEO Jensen Huang, even calling the maker of the world's most advanced computer chips during a Monday podcast interview. Nvidia has generally favored some chip exports to China as a way to build an American ecosystem for the technology and foster a more open relationship on AI, given the large number of researchers in China.

In both countries, AI's growing capabilities have alarmed senior officials and policymakers. The head of China's Ministry of State Security, its spy agency, warned in an article Sunday that AI poses numerous threats, from political and ideological security to cyberattacks. Chen Yixin argued that AI could endanger China's political and ideological security if used by those "with ulterior motives," singling out Anthropic's and OpenAI's powerful models as potentially able to "weaponize" cybersecurity vulnerabilities and threaten critical information infrastructure.

Other recent AI-related developments have also raised concerns in China about governance. These include a US security firm's discovery of AI vulnerabilities in WeChat, China's key messaging and payments app, according to Paul Triolo, a partner at DGA-Albright Stonebridge Group.

Although Chinese officials said after Trump's Beijing summit with Xi in mid-May that the two countries had agreed to pursue dialogue on AI development and governance, little progress has followed. Lizzi C. Lee, a fellow at the Asia Society Policy Institute's Center for China Analysis, said neither country "can simply seal itself off from the other," adding that both governments share an interest in setting guardrails for human control of AI. "We already know that neither government is going to abandon its competitive AI strategy or pursuit of technological advantage," she said.

Meanwhile, both nations are building AI alliances with like-minded countries. In July, China launched the World Artificial Intelligence Cooperation Organization, with 29 initial founding members including Russia and Pakistan, aimed at promoting global AI governance. This month, Xi called on BRICS members to accelerate the creation of a global AI governance framework and establish an open-source AI community. Some observers view the China-led grouping as a response to the US-led Pax Silica initiative, launched last year to strengthen collaboration on AI supply chains with countries such as Japan, the UK and Australia.

Asked about his expectations for the Trump-Xi meeting in an interview with CBS' "Face the Nation," Amodei said "the more long-term thing would be working together to put a speed limit on the rate of AI progress," though he acknowledged that would likely prove very difficult. In the run-up to the meeting, a social media post by Yuyuantantian, an account linked to Chinese state media, struck a similarly cautious note: "Only after the United States first demonstrates that the safety rules are equally effective for its own model companies can substantive discussions between the U.S. and China take place."