A quiet contradiction has emerged at the center of the debate over AI chip smuggling. Nvidia told a U.S. congressman that tracking its chips is technically possible, even as the industry group it belongs to publicly called such tracking potentially impossible. The gap between those two positions, reported by MIT Technology Review, is now drawing fresh scrutiny as a broader scandal unfolds around the company.

What Nvidia Said, and Didn’t Say Publicly

Representative Bill Foster, a Democrat from Illinois who co-sponsored the Chip Security Act, says Nvidia representatives told him directly that the chip-tracking measures in the bill were doable. The Chip Security Act would require advanced AI chips to have their locations verified, through facility inspections, company declarations, or a “ping-based” system where a chip automatically signals its location when it goes online, specifically to stop chips from being smuggled to China.

When Foster pressed Nvidia’s team on the technical side, their response, as he described it, was clear:

“Yeah, of course we can do this.”

Nvidia never put that acknowledgment in writing. Publicly, the company has stayed quiet. When Foster introduced the bill in May 2025, lawmakers said the technology existed. But in March 2026, the Semiconductor Industry Association, of which Nvidia is a member, formally opposed the bill and called its tracking requirements “new, untested, and potentially infeasible.” Nvidia did not respond to MIT Technology Review’s requests for comment on Foster’s account.

The Clue Hidden in a Product Nvidia Already Sells

There is evidence beyond Foster’s account that Nvidia can track chip locations. In December 2025, Nvidia launched a tool called Fleet Intelligence, designed to let customers monitor how their chips are performing across different data center locations. Nvidia confirmed to Reuters that the tool could indicate which country a chip is operating in. The company declined to say whether that same technology could be adapted to catch smuggling.

This matters because it suggests the core capability, knowing where a chip is, may already exist inside Nvidia’s own product lineup. The question is not whether the technology works, but whether Nvidia is willing to use it for export enforcement.

Why This Matters Beyond the Boardroom

For most people, chip smuggling sounds abstract. But the chips in question power the most advanced AI systems in the world. Controlling where they go is one of the main ways the U.S. government tries to limit the development of military AI by adversaries. If chips are slipping through to China, and there is an active investigation into a Singaporean firm called Megaspeed, which reportedly bought chips from an Nvidia packaging partner and is now under inquiry for smuggling, then that strategy has a real hole in it.

The scandal around Nvidia CEO Jensen Huang’s resignation and reported affair with Alice Huang, who runs Megaspeed, has pushed this technical debate into much sharper political relief. Foster put it plainly: there will be strong pressure on companies to demonstrate they know exactly where their chips are.

Even if the Chip Security Act passes as part of the Senate’s National Defense Authorization Act, the current version only tracks chip locations, it cannot remotely disable a chip found to be in the wrong hands. Foster says he plans to introduce new legislation next year that would require chips to hold a valid operating license before they can connect to a network, giving authorities an actual off switch.

What to Watch

Watch for whether the Chip Security Act gets included in the National Defense Authorization Act and what language survives. Keep an eye on the Megaspeed investigation and whether the Commerce Department inquiry, reportedly paused after Nvidia’s intervention, resumes. Foster’s promised new licensing bill, expected early next year, could significantly raise the stakes for every company that makes or sells advanced AI chips.