When President Trump and President Xi Jinping of China meet in Beijing this week, they are expected to discuss, for the first time, how to manage the risks of A.I. But in many ways, the two countries seem farther apart than ever on that question.<\/p>\n
Both the United States and China are racing to develop A.I.-powered weapons<\/a> that could wreak immense damage without human involvement. Powerful new A.I. models could enable cyberattacks that might cripple the world\u2019s banks and power grids. Experts have also raised the alarm about how A.I. could be misused by terrorists or even become sentient and wipe out humankind.<\/p>\n But the United States and China are locked in a battle for supremacy in A.I. that has left policymakers and some researchers increasingly wary of engagement, even as they warn of the technology\u2019s risks. In both countries, many policymakers fear that imposing guardrails on A.I. development \u2014 for example, curtailing the technology\u2019s ability to create bioweapons \u2014 would give the other country an opportunity to race ahead.<\/p>\n In recent weeks, U.S. officials have boasted that A.I. is helping them choose targets in the U.S. war in Iran. China, in a giant military parade last year, showed off drones<\/a> that were designed to fly autonomously alongside fighter jets.<\/p>\n An unmanned combat aerial vehicle at a parade for the Chinese military in Beijing\u2019s Tiananmen Square last year.Credit…Kevin Frayer\/Getty Images<\/p>\n Even those who favor cooperation must tread lightly for fear of political blowback, amid fraught geopolitical ties between the two countries.<\/p>\n For example, the organizers of NeurIPS, a major academic forum for A.I. development, announced<\/a> earlier this year that they would not accept papers from researchers at institutions subject to U.S. sanctions, including multiple major Chinese institutions and companies. The organizers eventually backtracked<\/a>, but many Chinese scholars said they would boycott the forum in protest anyway.<\/p>\n When Senator Bernie Sanders of Vermont brought Chinese and American A.I. researchers together for a public forum in Washington about how the countries could cooperate on preventing A.I.-related risks, he drew fierce criticism, including from Treasury Secretary Scott Bessent. Mr. Bessent, who is taking a leading role<\/a> in shaping American A.I. policy, wrote on X that Mr. Sanders should not be \u201cinviting foreign nationals to tell the United States how to regulate A.I.\u201d<\/p>\n The two governments have clashed openly over technology in the weeks before Mr. Trump\u2019s trip. The White House recently accused Chinese A.I. companies of \u201cindustrial-scale<\/a>\u201d theft of intellectual property from American companies. Beijing blocked Meta\u2019s attempted acquisition<\/a> of Manus, an A.I. start-up founded by Chinese engineers, citing national security concerns.<\/p>\n Treasury Secretary Scott Bessent is a taking a leading role in shaping American A.I. policy.Credit…Kenny Holston\/The New York Times<\/p>\n Some experts are cautiously optimistic that Mr. Trump\u2019s visit could kick-start renewed efforts to work together. The Trump administration, which previously had emphasized deregulating A.I., has in recent weeks expressed concern about the threat<\/a> that A.I. could pose to cybersecurity, after the U.S. company Anthropic announced its powerful new model, Mythos.<\/p>\n Cooperation on A.I. regulation is different from past efforts to manage global challenges. During the Cold War, the United States and the Soviet Union agreed on the need to avoid unauthorized or accidental use of nuclear weapons, so they shared information to reduce that threat.<\/p>\n With A.I., it is unclear exactly what risks the technology will pose or how best to mitigate them.<\/p>\n American scholars have generally sought to discuss existential risks \u2014 such as A.I.-designed pathogens or accidental nuclear war \u2014 and the possibility of super-intelligence, or artificial intelligence with capabilities that exceed the human brain. Chinese researchers and officials have more often highlighted risks to social stability and state security. That includes A.I.-fueled misinformation and chatbots producing answers that challenge the country\u2019s strict internet controls.<\/p>\n \u201cIt\u2019s already hard enough to get cooperation on a strategic technology,\u201d said Jeffrey Ding<\/a>, a professor at George Washington University who studies A.I. competition between the U.S. and China. \u201cIt\u2019s even harder when we don\u2019t necessarily have a clear and harmonized view on what problem we\u2019re trying to tackle.\u201d<\/p>\n Those diverging priorities have dogged efforts at cooperation from the start. In November 2023, then-President Joe Biden and Mr. Xi announced an agreement to discuss A.I.-related risks \u2014 the first time the countries\u2019 leaders had publicly acknowledged the topic. About six months later, representatives from both sides met in Geneva. American officials wanted to discuss the possible misuse<\/a> of A.I., including by China. Chinese officials instead focused on U.S. export controls that they thought were designed to hamper China\u2019s A.I. development.<\/p>\n Officials had initially described that meeting as the first of multiple. No more materialized.<\/p>\n In late 2024, Mr. Xi and Mr. Biden, in their last meeting of Mr. Biden\u2019s presidency, did agree to ensure human control over nuclear weapons<\/a>, rather than leaving it to artificial intelligence. But there have been no more details on that issue, either.<\/p>\n The United States was not worried, at first, about including China in A.I. safety discussions partly because it was confident of its technological lead, said Scott Singer<\/a>, who studies China\u2019s A.I. policy at the Carnegie Endowment for International Peace. It wasn\u2019t until early 2025, with the arrival of DeepSeek, the Chinese start-up with its own powerful A.I. model, that American policymakers fully recognized that Chinese A.I. could affect U.S. national security, too, he said.<\/p>\n DeepSeek headquarters in Hangzhou, China, earlier this year.Credit…Jade Gao\/Agence France-Presse \u2014 Getty Images<\/p>\n That recognition has hardened Washington\u2019s competitive mind-set. The Trump administration has explicitly named<\/a> \u201cwinning the race\u201d against China as a reason to loosen regulations on A.I.<\/p>\n Still, even as formal talks on A.I. have stalled, scholars from both countries have held informal discussions, often through academic conferences or think tanks. Participants described these meetings as vibrant, and said they have produced many suggestions for cooperation, such as an emergency hotline in case of an A.I.-related accident, or shared standards for testing whether A.I. has the ability to synthesize biohazards.<\/p>\n But these conversations have not been immune from political pressures.<\/p>\n