Earlier this month, OpenAI chief executive Sam Altman pitched a new way President Donald Trump might win the Nobel Peace Prize he has long coveted: reaching a bargain with China on safeguards for artificial intelligence.
“Even if just the U.S. and China could agree on some shared standards and testing for development of this technology, I think that’d be a wonderful accomplishment that the two of them can deliver,” Altman told Fortune magazine.
But although Altman and other American tech executives are set to dine with Trump and Chinese President Xi Jinping on Thursday as the leaders meet in Washington, the White House has not signaled any interest in a comprehensive deal.
The presence of the top executives at the dinner reflects the increasingly prominent position that AI safety has taken on the agenda at a time when Washington has been gripped by the danger that out-of-control AI could pose to the human race.
As cutting-edge AI systems become more powerful, some of the industry’s leading figures have become increasingly worried about their technology’s potential to cause catastrophic harm. They see a need for coordination between China and the United States, home to the top companies building the technology, potentially including a deal on slowing the pace of development.
But Washington and Beijing have starkly different views regarding the biggest risks posed by AI. And the Trump administration remains divided over how to oversee the technology, with the president favoring minimal new regulation even as some of his aides push for stronger measures.
That leaves experts seeing few prospects for a meaningful breakthrough between the two countries.
“The bottom line is there is no mutual trust,” said Ryan Fedasiuk, a fellow at the American Enterprise Institute. “No grand bargain. No agreements to limit either side’s capability development.”
Despite how the risks of AI have grabbed attention in recent weeks, there are few signs that the two countries’ positions have significantly shifted, Fedasiuk said.
On Tuesday, Trump said in an address to the U.N. General Assembly that the U.S. remained ahead in the development of AI, which he said he was rebranding as “superintelligence,” and that he had no interest in using the government’s power to slow it down.
“We’re leading now over China by a lot, and everyone else, and we’re going to keep it that way,” he said. “We’re going to keep it very, very straight and very strong. I’m not going to stifle growth of something that will be bigger than the industrial revolution.”
The White House expects Trump will discuss AI with Xi, as well as China’s 2025 agreement to pause restrictions on the exports of rare earth minerals needed to build computer chips, said a White House official, speaking on the condition of anonymity to describe private plans. The official declined, however, to provide many details about the White House’s positions until after the leaders talked.
Thursday’s dinner, scheduled to be held in the White House’s East Room, will include SpaceX founder Elon Musk, who like Altman, has endorsed slowing down the pace of AI development, according to the White House official.
But also on the invitation list is Jensen Huang, the chief executive of chipmaker Nvidia, said the official. Huang has not joined the call for slowing down, saying the warnings of dire risks from AI are overblown — a topic he discussed with Trump, who called him in the middle of a tech conference appearance last week.
Other tech figures, including Amazon founder Jeff Bezos, Google CEO Sundar Pichai, Dell Technologies CEO Michael Dell and Apple chairman Tim Cook, are expected to be there, the White House official said. So, too, are Meta CEO Mark Zuckerberg, OpenAI president Greg Brockman and Microsoft CEO Satya Nadella, according to the official and people familiar with their plans. (Bezos owns The Washington Post.)
Senior Trump officials have emphasized that the U.S. is in a race to dominate the development of AI technology and have accused Chinese companies of effectively stealing from American firms using a technique called distillation, which allows developers to perform the same complex tasks with fewer computing resources.
But the administration also has been cautious about imposing any limits on the kind of open AI models largely produced in China, which are popular with businesses because they are often more affordable and can be modified by outside developers. The administration has also approved the export of advanced computer chips that some experts say are fueling China’s ability to keep up.
Democrats have argued that those moves have undermined the president’s warnings that the U.S. needs to maintain its competitive edge over China through a light-touch regulatory approach.
In a publicletter to Trump, Sen. Elizabeth Warren (D-Massachusetts) accused the administration of advancing the interests of tech companies such as Nvidia over those of the American people.
“Your decision to allow the export of chips to China puts our national security at grave risk. China is working feverishly to develop advanced AI, including AI weapons systems that can be turned against the United States,” Warren wrote. “… The American people expect you to press President Xi about the risks of AI and work with China and countries around the world to set rules and regulations for this technology.”
For now, the two countries appear to have agreed to continue talking, but little else.
On Sunday, Treasury Secretary Scott Bessent — one of the Trump administration’s leading players on AI — met Chinese Vice Premier He Lifeng at JPMorgan’s headquarters in New York to discuss trade and AI.
Bessent said he proposed a “notification mechanism” for the two nations to alert each other about AI incidents with national security implications, though the Chinese have not publicly agreed to participate.
The two countries “want a shared vision of common goals and common threats,” Bessent told reporters following the meeting.
“We think that, just like any cross-border activity, that moving from more opaque to more transparency between the number one and number two AI powers in the world is very important.”
American and Chinese officials plan to continue formal discussions in November in Shenzhen, China, Bessent told CNBC on Monday.
For AI industry leaders worried about the dangers presented by the most powerful forms of the technology, managing China is key. Chinese companies have tended to release their AI models openly, allowing others to use and modify them freely. That means that even if American companies were to lock down their technology, there could still be significant risks if Chinese models could readily be used to carry out cyberattacks or help develop biological weapons, two of the bigger risks flagged by those who warn of AI’s potential for harm.
In an influential essay calling for a slowdown in the development of AI, Anthropic chief executive Dario Amodei laid out a two-pronged approach to China. He urged cutting off China’s access to top-end American AI chips and stopping the country’s companies from stealing U.S. innovations. Then, with American leadership assured, Amodei believes it would be possible to cooperate with China, he wrote.
“We must not be naïve here: the geopolitical stakes are so high that there will likely be stark limits on what can be achieved, especially at first,” he wrote.
The post AI leaders say U.S. and China need a deal. That probably won’t happen. appeared first on Washington Post.
