Bernie Sanders

Senator Bernie Sanders wants to prevent the “runaway train” of unregulated development of artificial intelligence by working with China.

The Vermont independent has a point, as anyone who has watched the Terminator movies knows, but cooperation with the Chinese state would nonetheless not be in America’s interest.

Sanders is passionate about the issue. “What I don’t understand is how anyone can ignore when the world’s leading AI scientists tell us that, as AI becomes smarter than humans, it could operate independently of human control—with possible catastrophic results,” he wrote on X on April 30.

The day before at the U.S. Capitol, Sanders hosted American and Chinese scientists who advocated cooperation. “If President [Donald] Trump can sit down with [Chinese] President Xi [Jinping], I think leading scientists in America can sit down with leading scientists in China to discuss how in fact we can work together,” he told reporters.

Trump may discuss AI issues with Xi when they meet in Beijing this week. If so, this will not be the first time an American leader has tried to start a conversation on the topic.

President Joe Biden raised the issue of AI cooperation with Xi when the two leaders met in California in November 2023. Chinese officials in May 2024, however, rejected talks with America on limiting the use of artificial intelligence for controlling the launch of nuclear weapons.

“Nobody wants to see AI-controlled nuclear weapons, right?” asked Joe Wang, a former State Department and National Security Council (NSC) staffer at the Arlington, Virginia-based Special Competitive Studies Project, which specializes in AI and emerging technologies. “Like, even the craziest dictator can probably agree.”

Whether crazy or not, Xi is not willing to talk about the topic.

Perhaps that unwillingness is fortunate for the United States. Yes, no American official wants China’s machines to have the ability to trigger Armageddon, but that does not mean cooperation is possible.

Or in America’s interest. “Bernie Sanders’s neo-Luddite stance on AI and his urging that the U.S. capitulate to China by acceding to CCP blandishments on AI development ‘cooperation’ amount to de facto calling for America’s surrender in the hybrid war that Xi Jinping is waging for global supremacy,” Charles Burton of the Sinopsis think tank told me this month. “Beijing must be rejoicing at Sanders’s wrong-headed naivety.”

In the race for AI supremacy—sometimes characterized as the race for the 21st century—the U.S. holds a lead. Cooperation would tend to erode that edge as water runs downhill. “Of course, Beijing is growing increasingly concerned that the Trump administration has lit the afterburners of innovation, setting the U.S. to hold the lead in AI,” Blaine Holt, a retired U.S. Air Force brigadier general who comments on China and other matters for Newsmax, told me, in response to Sanders.

“Panic must be setting in because the Communist Party’s system is not an incubator that nurtures new technologies like the U.S.,” Holt added. “So China must turn to its ‘friends’ in D.C. to hold back America. Maybe Sanders would be so kind to describe why compromising a competitive advantage is in our national interest.”

Sanders and Representative Alexandria Ocasio-Cortez (D-NY) unveiled their Artificial Intelligence Data Center Moratorium Act, which would prohibit the construction or upgrading of new AI data centers until Congress enacted a regulatory framework. Sanders office, as Fox News reported, said the proposed law was designed to “slow down the development of AI.”

The Vermont senator, who has pointed to Cold War-era arms control agreements as models, wants a global pact on AI.

Sanders needs to pick a better example. Beijing has repeatedly rejected Washington’s attempts to discuss nuclear arms control, and China has—despite its obligations in the global Nuclear Non-Proliferation Treaty—spread nuclear weapons technology, equipment and materials to, among others, Pakistan, Iran and North Korea. America, on the other hand, prevented friends and allies from developing such weapons.

Moreover, a global AI pact would be completely unworkable. For one thing, it would be hard for a free society to ensure compliance as AI development is not something that can be monitored like warheads on top of missiles.

An agreement would be difficult to enforce against others as verification would be virtually impossible for the same reason.

There is no question that AI, as Sanders says, could pose an “existential threat,” but an agreement with China is not the answer as there is no evidence that Xi Jinping is more concerned about the fate of the human race than that of the Chinese he rules. So, perhaps he is more dangerous than the “craziest dictator.”

Is there any solution?

Perhaps the only one is to let Americans innovate, dominate AI and trust the good instincts of their tech community to develop sensible and humane standards. Such standards are unlikely to emerge from a world dominated by the researchers of a militant Chinese Communist Party.