Nvidia has privately indicated that it could support the kind of location tracking lawmakers want to impose on advanced AI chips, creating a sharp divide between the company’s behind-the-scenes technical discussions and the semiconductor industry’s public objections. Representative Bill Foster of Illinois said Nvidia representatives told him the requirements in his proposed Chip Security Act could be implemented, even though the Semiconductor Industry Association has called the mandates “new, untested, and potentially infeasible.”
Foster, a Democrat on the House Committee on Science, Space, and Technology, introduced the bill with Representative Bill Huizenga in 2025. The legislation would require companies handling advanced AI processors to verify where those chips are located, with possible methods including facility inspections, corporate attestations, or electronic signals that identify a chip’s location. The bill is among several China-focused technology measures receiving renewed attention as lawmakers assemble provisions for the Senate’s National Defense Authorization Act.
The apparent feasibility of the proposal is partly illustrated by Nvidia’s own software. In December 2025, the company announced an opt-in fleet-management service that gives data-center operators a dashboard for monitoring GPU usage, errors, power consumption, thermal conditions, and other operational data across distributed facilities. Nvidia says the service uses a customer-installed software agent to transmit GPU telemetry to its NGC cloud platform and organize systems by physical or cloud-based compute zones. Nvidia’s description of the service does not present it as an anti-smuggling system, but its ability to provide visibility across global GPU fleets helps explain why lawmakers believe location verification is technically achievable.
The political pressure comes as investigators examine alleged routes used to divert Nvidia-powered servers toward China. Singapore-based Megaspeed, associated with Alice Huang, has reportedly drawn scrutiny over suspected semiconductor smuggling. Supermicro, which packages Nvidia processors into server systems, has also faced investigations. Federal prosecutors charged Supermicro executive Yih-Shyan “Wally” Liaw and two others in March over an alleged scheme involving roughly $2.5 billion in servers containing Nvidia H200 and B200 GPUs routed through Southeast Asia, according to coverage of the case.
Tracking, however, would not automatically stop a shipment once a chip left an approved facility. The legislation’s most aggressive option would use recurring electronic “pings” to establish where a processor is operating, potentially allowing authorities to flag a device that appears in an unauthorized country. That approach also raises practical questions about privacy, connectivity, tampering, ownership, and whether location data could be trusted across cloud providers and international data centers.
Nvidia has not publicly endorsed Foster’s characterization of the private conversations, and the company has declined to say whether its fleet software could be adapted to prevent diversion. Its public position has emphasized that illegally redirected controlled products represent only a small share of the computing capacity already available in China. Still, if Congress advances the Chip Security Act, Nvidia’s existing monitoring infrastructure—and its reported private assurance that the proposed safeguards are workable—will likely become central to the debate over whether advanced GPUs should be treated less like ordinary hardware and more like controlled strategic technology.
