Representatives from Nvidia told Illinois representative Bill Foster that it would be technically possible to implement the anti-smuggling measures called for in the Chip Security Act, Foster told MIT Technology Review in an interview on Thursday, even while an industry group claimed they were “potentially infeasible.”  Foster, a Democrat who serves on the Committee on…

Nvidia representatives have privately told Rep. Bill Foster (D-Ill.) that the company could meet proposed federal requirements to verify where advanced AI chips are being used—an assurance that contrasts with public warnings from the semiconductor industry that such measures may not be workable.

In an interview with MIT Technology Review on Thursday, Foster said Nvidia staff involved in technical conversations about his proposed Chip Security Act indicated the tracking approach envisioned by the legislation is within the company’s capabilities. “They never put anything in writing,” Foster said, but added that the tenor of the discussions was effectively: “yeah, of course we can do this.”

Foster, who serves on the House Committee on Science, Space, and Technology, is a co-sponsor of the Chip Security Act, introduced in May 2025. The bill aims to curb the diversion of cutting-edge US AI processors—particularly graphics processing units (GPUs)—by requiring mechanisms that can validate the geographic location of advanced chips and help deter smuggling to China.

The proposal has gained renewed attention as Congress considers multiple chip-security measures for possible inclusion in the Senate’s National Defense Authorization Act. It is one of three bills now being weighed, amid heightened scrutiny following turmoil surrounding Nvidia CEO Jensen Huang, including his resignation and reports of an affair involving an alleged Chinese spy.

When lawmakers rolled out the Chip Security Act last year, they argued that the underlying technology already exists, though Nvidia has not publicly endorsed that claim. In March 2026, the Semiconductor Industry Association (SIA)—a trade group that includes Nvidia as a member—formally objected to the legislation. In its opposition, the SIA described the bill’s tracking mandates as “new, untested, and potentially infeasible.”

Foster’s account suggests Nvidia’s private posture is more accommodating than the industry’s public line. Evidence from Nvidia’s own product roadmap also points to technical pathways for location verification. In December 2025, Nvidia introduced “Fleet Intelligence,” a tool designed to help customers monitor how Nvidia’s most advanced chips perform across data centers. Nvidia confirmed to Reuters that the system could reveal the country where chips are operating, but the company has not said whether it could be adapted to discourage or detect illicit diversion.

Nvidia did not respond to requests for comment on Foster’s characterization. When MIT Technology Review asked the company earlier in October whether Fleet Intelligence could be used in anti-smuggling efforts—an area of growing concern among lawmakers—Nvidia declined to answer directly. Instead, it provided a statement asserting that diversion of controlled US products is minimal compared with China’s existing domestic computing capacity: “Governments and media have found insignificant diversion of controlled U.S. products—a drop in the bucket compared to the ocean of domestic compute China already has.”

The debate is unfolding against a backdrop of ongoing allegations involving firms accused of moving restricted GPUs into China. Alice Huang—reported to have had a relationship with Jensen Huang—runs Megaspeed, a Singapore-based company under investigation for allegedly smuggling high-end GPUs to China. A US Commerce Department inquiry into Megaspeed was reportedly stopped after Nvidia intervened. Reports also say Megaspeed purchased chips from Supermicro, which integrates Nvidia chips into server systems.

As drafted, the Chip Security Act would allow several compliance routes: physical inspections of facilities, company attestations, or “ping”-based location checks. The ping approach is seen as the hardest to evade because it would require a GPU to transmit location information when it comes online. In September, the Center for Security and Emerging Technology concluded that ping-based tracking would likely be the most cost-effective option.

Even if enacted, the bill’s framework would focus on identifying where chips are operating—not on remotely disabling them if they are found to have been diverted, Foster noted.

Why it matters

Advanced AI chips are strategic assets, and US officials worry that smuggling undermines export controls meant to slow China’s military and surveillance capabilities. If Nvidia can technically support location verification, pressure may intensify for industry to adopt tracking safeguards rather than resist them as impractical. The next key signal will be whether any chip-security provisions, including the Chip Security Act, make it into the NDAA—and how enforcement would work without a “kill switch.”