China Bypasses US Chip Restrictions as Blacklisted Universities Continue to Access Nvidia AI Processors
Blacklisted Chinese Universities Continue to Obtain Restricted Nvidia AI Chips Despite US Export Ban
US efforts to prevent China from accessing advanced artificial intelligence technology are facing fresh questions after at least six Chinese universities on Washington’s trade blacklist reportedly obtained restricted Nvidia processors in recent years. According to a CNN investigation, procurement documents and business records suggest that American export restrictions have not completely prevented sensitive AI hardware from reaching institutions facing US sanctions.
The findings are particularly concerning because two of the universities, which reportedly have close links to China’s defence establishment, purchased advanced processors within the past few months. These developments raise concerns about the effectiveness of restrictions designed to limit China’s access to technologies that could support military research and advanced AI development.
Third-Party Suppliers Help Chinese Institutions Access Restricted Chips
The US Commerce Department began imposing increasingly stringent restrictions on advanced semiconductor exports to China in late 2022. The measures target high-performance processors, including Nvidia’s A100 and H100 series, which are widely used for training sophisticated AI models.
Universities and organisations included on the US Entity List face additional restrictions on obtaining American technology without special government authorisation.
However, procurement records suggest that some Chinese institutions are obtaining restricted hardware through indirect purchasing arrangements. Instead of placing orders directly with Nvidia, buyers reportedly turn to Chinese technology distributors, server manufacturers and smaller system integrators.
These intermediaries can source equipment through complicated supply networks involving third-party distributors across different Asian markets, making it harder to establish the final destination of individual processors.
Defence-Linked Universities Raise Security Concerns
Some of the institutions identified in the investigation are associated with China’s defence research network. Their academic and research activities include areas such as hypersonic technology, military simulations and autonomous defence systems.
Advanced graphics processing units (GPUs) are essential for developing and training large AI models. Nvidia processors are particularly sought after because of their computing capabilities and established software ecosystem.
Even limited quantities of restricted chips can provide significant computing resources to research laboratories. When installed in servers or upgraded workstations, these processors can support research with both civilian and potential military applications.
Nvidia Maintains Compliance With US Regulations
Nvidia has consistently stated that it follows applicable US export-control requirements and expects its distributors and business partners to comply with the law. The company has also maintained that it does not directly sell restricted products to organisations prohibited from purchasing them.
However, monitoring hardware after it enters secondary distribution channels remains a major challenge. Once servers and processors are resold through multiple intermediaries, identifying their ultimate users and tracking their movement across international borders becomes considerably more difficult.
Washington Faces Challenges in Enforcing Chip Restrictions
The reported purchases highlight the difficulties facing US policymakers as they attempt to restrict China’s access to advanced semiconductor technology.
Although Washington has expanded export controls and increased scrutiny of technology suppliers, strong demand for powerful AI processors has created lucrative opportunities for unauthorised distributors and grey-market traders.
The latest findings raise questions about whether tightening restrictions on chip manufacturers alone can prevent sensitive technology from reaching prohibited buyers.
As competition between the United States and China intensifies in artificial intelligence and semiconductor development, controlling the global supply chain for advanced processors is likely to remain a major strategic and regulatory challenge.
Comments are closed.