A proposed system to track advanced artificial intelligence chips worldwide would create two major cybersecurity vulnerabilities that could be exploited by hackers ranging from individual activists to state-sponsored military units, according to a new analysis published by the R Street Institute. The commentary examines the Chip Security Act, currently moving through Congress as part of the National Defense Authorization Act, which would mandate ping-based location verification to monitor semiconductors subject to export controls. The analysis warns that this tracking approach could undermine the very security it's meant to protect.
The tracking system would rely on modifying the Trusted Execution Environment—a secure enclave in AI chips designed to protect data from compromised operating systems—and a global network of 39 to 500 landmark servers that estimate chip location based on ping response times. The best-case accuracy for this location estimation carries a 19-mile error radius, meaning chips located within that distance of each other can't be distinguished. Each advanced AI chip carries a value in the tens of thousands of dollars, while a single rack costs over $5 million, making them highly attractive targets for well-funded adversaries. The system would need to track semiconductors already deployed in the field, not just new production.
The report identifies how both components present exploitable weaknesses. Adding a network pathway to the Trusted Execution Environment would "invalidate the purpose of the TEE altogether" by introducing vulnerabilities to data confidentiality and integrity, the analysis states. The landmark server infrastructure could fall victim to Distributed Denial of Service attacks that would cause "the entire location system to go blind, simultaneously making every tracked chip noncompliant," according to the commentary. Countries subject to export controls that control their own network infrastructure could manipulate network traffic to make chips appear to be in different locations, while covered entities could simply build data centers within the 19-mile uncertainty zone near nations not on the export-control list—a situation that already exists between Chinese-run data centers in Johor, Malaysia, and Singapore, just 10 miles apart.
The strategic logic behind the system faces a fundamental contradiction, the report argues. Even the feasibility study supporting the Chip Security Act acknowledges that the ping-based location verification system is "unlikely to withstand advanced persistent threat actors"—a term originally coined to describe state-sponsored Chinese hacking operations. Building a chip-tracking mechanism to prevent Chinese diversion efforts that can't survive Chinese cyberattacks represents an approach "entirely devoid of strategic merit," the analysis concludes. Beyond nation-state actors, the system's vulnerability to common Distributed Denial of Service attacks means less sophisticated hackers could also compromise it. The commentary warns this export-control measure is moving forward without the level of debate needed to properly evaluate its consequences, noting that policymakers may next turn to mandating remote kill switches in advanced semiconductors when they discover location tracking lacks an enforcement mechanism.

