Nvidia privately said it can track AI chips it publicly calls untrackable

Nvidia representatives privately told Illinois Representative Bill Foster that tracking advanced AI chips to prevent smuggling to China is technically possible — even as the company's own industry group publicly called the idea "potentially infeasible."

Foster, a Democrat on the House Committee on Science, Space, and Technology, revealed the private acknowledgment in an interview with MIT Technology Review on Thursday. He is a cosponsor of the Chip Security Act, a bill that would require geotracking of advanced AI chips to verify their location and block their diversion to China.

The gap between what Nvidia says in private and what its allies say in public now sits at the center of one of the most consequential fights in US export policy. The stakes are enormous: advanced GPUs are the raw material of AI development, and China's access to them through smuggling channels undermines the entire architecture of US export controls.

What did Nvidia actually say?

"They never put anything in writing," Foster told MIT Technology Review. But during technical discussions about the safeguards, he said, the representatives' reaction was "yeah, of course we can do this."

That private admission goes further than anything Nvidia has said publicly. When the Chip Security Act was introduced in May 2025, lawmakers asserted the tracking technology already existed. Nvidia never confirmed it.

In March 2026, the Semiconductor Industry Association — which counts Nvidia as a member — explicitly opposed the bill. The group wrote that its mandates for chip tracking were "new, untested, and potentially infeasible."

Nvidia did not respond to requests for comment or confirmation of Foster's characterization.

Does Nvidia already have the technology?

Evidence suggests it does. In December 2025, Nvidia announced a tool called Fleet Intelligence. It lets customers monitor how advanced chips perform across different data center locations. Nvidia confirmed to Reuters that the tool could indicate which country chips were in — but the company would not elaborate on whether the technology could be used to prevent smuggling.

In early October, as lawmakers grew increasingly concerned about Nvidia's role in smuggling cases, MIT Technology Review asked the company directly whether Fleet Intelligence could stop the practice. Nvidia declined to answer directly and instead sent a statement: "Governments and media have found insignificant diversion of controlled US products—a drop in the bucket compared to the ocean of domestic compute China already has."

Foster said he never asked Nvidia's representatives a blunt follow-up question: if the company can technically track its chips, why isn't it doing so voluntarily?

Why the smuggling scandal is pushing the bill forward

Three chip-security bills are now under consideration for inclusion in the Senate's National Defense Authorization Act. Each has drawn renewed attention following the resignation of Nvidia CEO Jensen Huang and reports of his affair with an alleged Chinese spy.

The scandal's cast of characters keeps expanding:

• Alice Huang, with whom Jensen Huang reportedly had an affair, runs Megaspeed, a Singaporean firm now under investigation for smuggling advanced GPUs to China.
• A US Commerce Department inquiry into Megaspeed was reportedly halted after intervention by Nvidia, according to the New York Times.
• Megaspeed reportedly bought chips from Supermicro, a company that packages Nvidia's chips onto servers.

This web of alleged smuggling has changed the political math on Capitol Hill. Foster says the controversy has fueled support among his colleagues for the bill's rules.

"There will be a lot of political pressure, in the wake of everything that's happened, on the companies to actually do a very good job of the geographic validation of where the chip actually is," Foster told MIT Technology Review.

How would chip tracking actually work?

The Chip Security Act as drafted would require a mechanism to track chip locations through one of three routes:

• Inspections of facilities where chips are installed
• Attestations by companies operating the hardware
• Ping-based location mechanisms, where the GPU sends a signal revealing its location whenever it comes online

The ping-based approach is the hardest to circumvent, and it has independent backing. In September, the Center for Security and Emerging Technology found it would be the most cost-effective option of the three.

What happens after the bill passes?

Even if the Chip Security Act passes as part of the NDAA, it has a built-in limitation. The tracking mechanisms would tell investigators where smuggled chips are — but they would offer no way to disconnect them once diversion is confirmed.

Foster intends to fix that. He told MIT Technology Review he plans to introduce stricter legislation in the new year. The proposed rule would require licenses to operate advanced GPUs. Before a GPU could connect, its valid license would need to be verified.

That approach would flip the enforcement model. Instead of finding smuggled hardware after the fact, the hardware itself would refuse to run without a valid, verified license — turning every advanced GPU into its own customs checkpoint.

For Nvidia, the political ground is shifting quickly. A company that spent 2026 publicly minimizing diversion as "a drop in the bucket" now faces a private admission, reported by a sitting lawmaker, that it could build the safeguards it has resisted — and if the NDAA carries the Chip Security Act through, it may soon have no choice.