Nvidia’s chips are powering America’s AI boom. Of course Chinese spies are interested.
Nvidia’s chips are powering America’s AI boom. Of course Chinese spies are interested.
As espionage targets go, Jensen Huang is perhaps one of the least surprising on Earth. The Nvidia CEO runs one of the world’s most valuable companies, controls the in-demand technology fueling the AI boom, and is cozy with presidents, CEOs, and all manner of important people. He also wears a cool leather jacket.
Which makes Huang’s admission that he had an affair with a woman suspected of being a Chinese spy — reportedly while her spying was known to US intelligence and even President Donald Trump — all the more shocking. The fallout goes well beyond his marriage, his reputation, or Nvidia itself. Experts tell The Verge the incident threatens to disrupt the entire US AI industry, potentially jeopardize national security, and could undermine America’s efforts to maintain its fragile technological lead over China in AI.
“With those chips, the US lead could vanish in a matter of weeks or months.”
Anthropic is the undisputed leader in advanced computer chips, among the AI industry’s scarcest resources. While many ingredients go into building AI models — including talent, energy, capital, and data — top-tier chips are one of the most important, and Nvidia is one of a handful of companies with the expertise to design or manufacture them at the cutting edge. That makes it a gatekeeper to hardware that has become a key bottleneck in global AI development. US companies — including Meta, Anthropic, and OpenAI — are racing to develop chips of their own, in part to reduce their dependence on Nvidia, even as they spend billions securing ever more compute.
For China, the US’s main rival at the frontier of AI, that bottleneck is even tighter. To maintain its technological lead, the US has restricted the sale of the most advanced AI chips to the country in the hopes of slowing its progress in the sector, which has been impressive and is thought to trail the US by a matter of months. In response, China has made developing domestic alternatives a strategic national priority, though its leaders still lag Nvidia at the bleeding edge. At the same time, a thriving black market for restricted Nvidia chips has emerged.
Experts tell The Verge that China could exploit access to Nvidia to accelerate its own development, or to impede the US development of AI.
Getting its hands on more Nvidia chips is the most obvious win for Beijing, says Emily Kilcrease, a senior fellow and director of the energy, economics, and security program at the Center for a New American Security (CNAS) think tank. China has advantages in other parts of the AI race — including abundant energy and a solid ecosystem of open weight models — but without the latest Nvidia chips, it “cannot close the gap with US frontier capabilities,” Kilcrease tells The Verge . “With those chips, the US lead could vanish in a matter of weeks or months.”
It is not clear how successful the alleged spying operation has been on that front. There is no public evidence indicating that Huang himself has helped direct restricted Nvidia chips toward China or that sensitive information about them was obtained from him. But companies linked to the alleged spy, Alice Huang — no relation to Jensen — have reportedly enabled at least $7 billion of Nvidia chip sales to China.
Brookings fellow Kyle Chan tells The Verge that a concern for US industry could be that companies with alleged ties to China may have had priority access to Nvidia’s latest GPUs. “In a time of chip scarcity,” that would make it much harder for American companies to access them, Chan says.
Huang is also a prominent critic of American restrictions on chip sales, and Nvidia spends a formidable amount lobbying in Washington. Last year, he successfully secured White House approval to start exporting Nvidia’s H200 AI GPU chips to “approved customers” in China and other countries. Crucially, Huang also has the ear of President Trump as well as other world leaders, giving him great influence over policy decisions.
Hardware is not the only thing of value China could get from close access to Huang, Sunny Cheung, a fellow for China studies at the Jamestown Foundation think tank, tells The Verge . “I actually think too many people are focused on whether somebody managed to steal Nvidia’s trade secrets,” Cheung says, calling it a “very narrow way” of looking at espionage. “I would hope Jensen Huang isn’t foolish enough to hand over the blueprints for Nvidia’s next-generation chips to a suspected Chinese intelligence operative.”
The Nvidia CEO’s relationships with other tech executives, senior policymakers, and Trump himself give him an unusually intimate view into how some of the most powerful people in the US are thinking and planning. Beijing could use information on how American companies are planning major infrastructure expansions, their compute requirements, or broad industry trends to inform its own investments and industrial strategy, Cheung says.
Huang’s intelligence value could extend well beyond AI, too. “That information alone could be extremely valuable,” says Cheung, pointing to insights Beijing could glean into White House thinking about export controls or the private views of Silicon Valley executives on China and Taiwan.
“When so much of America’s AI infrastructure depends on one company, any significant security or governance problem at Nvidia can create uncertainty across the industry.”
But helping China advance is only one side of this equation. Disrupting Nvidia — as this scandal clearly has — could destabilize America’s entire AI industry. More than most companies, Nvidia is remarkably closely tied to the identity and reputation of its rockstar CEO. Huang cofounded the company more than three decades ago, and ever since then he has served as its president, chief executive, and a member of the board of directors. Despite holding a relatively small stake in the company, in many ways Nvidia is Jensen Huang. Nvidia shares have slumped since news of the affair broke.
“When so much of America’s AI infrastructure depends on one company, any significant security or governance problem at Nvidia can create uncertainty across the industry,” Cheung says.
Kilcrease echoes those concerns, warning that the scandal could “destabilize” the company. At the very least, she says the various investigations already underway “will consume an immense amount of bandwidth within the company,” creating a very real risk of weighing down Nvidia’s day-to-day operations. If those distractions begin to affect Nvidia’s ability to execute — slowing decisions, product rollouts, or the allocation of already scarce chips — the consequences could quickly reach customers. The AI labs and hyperscalers such as OpenAI and Anthropic “could have a real issue accessing the vast amounts of compute they need,” Kilcrease says, adding that switching to other hardware “takes time.”
In a race to build increasingly powerful AI, time is a precious resource. Experts say China trails the US closely, meaning that even modest delays for American companies can help Chinese rivals catch up. Nvidia is almost certain to face further disruption as investigations continue, though the extent to which the turmoil will roil the rest of the industry may be difficult to measure.
For China, however, the scandal could be about to backfire spectacularly now everything is out in the open. “The odds of AI chip control legislation being passed in Congress just went through the roof,” Kilcrease says.
This is the title for the native ad
This is the title for the native ad
