U.S. export controls on AI chips are being circumvented as Chinese firms like ByteDance access powerful NVIDIA GPUs remotely through data centers in countries like Malaysia. This loophole, combined with complex corporate shell structures, allows them to train frontier models, rendering the current import-focused restrictions largely ineffective.

The most potent criticism of the U.S. chip controls wasn't flawed strategy, but the chronic underfunding and limited capacity of agencies like the Bureau of Industry and Security (BIS) to effectively enforce complex export bans against determined adversaries.

The most significant sanctions loophole isn't physical chip smuggling but 'compute smuggling.' Chinese firms establish shell companies to build and operate data centers in neutral countries like Malaysia. They then access this cutting-edge compute power remotely, completely bypassing physical import restrictions on advanced hardware.

The Trump administration's failure to replace the Biden-era "AI diffusion rule" for over a year after declaring it non-enforced is the direct cause of a loophole allowing Chinese subsidiaries to buy advanced chips. This was not a policy choice but a significant bureaucratic failure that undermined the entire export control regime.

The focus of US chip export controls is moving downstream from individual chips to complete server systems. This makes enforcement a complex issue of customs, forged documents, and international logistics channels, as seen in the Taiwanese investigation into NVIDIA-powered server smuggling.

The current US strategy is contradictory. While taking extreme measures to block allies like Canada from accessing advanced US AI models, the administration's inaction has left open loopholes that allow Chinese firms to freely acquire the very chips needed to build competing models. This highlights a critical disconnect.

The Commerce Department's export control order against Fable may lack legal authority. Existing laws and the department's own guidance explicitly state that export controls do not cover cloud services or software-as-a-service. This makes the "ban" legally tenuous and vulnerable to a court challenge.

The US strategy for controlling AI chip exports has evolved from blocking product sales to supervising entire networks. Authorities now focus on loopholes like foreign subsidiaries, third-country routing, and cloud access, signaling a more sophisticated approach to compute governance.

Evasion tactics for U.S. chip controls have grown more sophisticated than simple remote access. Chinese companies are now building entire data centers in Southeast Asian countries like Malaysia and using complex shell companies and financial arrangements to obscure their identities when purchasing compute resources, making enforcement significantly harder.