China AI Export Controls Take Effect; Huang Calls US Controls a Failure

China's interim AI export control catalogue took effect October 16, 2026, covering 40 technologies across AI, biotech, and solar sectors, with a 30-day public consultation running through November 15 before the rule is finalized. For AI specifically, 14 technologies face outright export bans and 26 face restrictions requiring government approval, covering large models, AI agents, computing clusters, and AI-assisted vulnerability discovery; APIs, apps, SaaS products, and already-public open-source technologies are exempt. Dean Ball characterized the move as giving the Chinese government optionality to shut down future open-weight model exports at will, and described it as mirroring the US model-weight export control regime. The final catalogue is expected to narrow vague categories like 'large models' and 'intelligent agents' into specific, measurable thresholds.

Nvidia CEO Jensen Huang has argued that US export controls on advanced chips to China have been a failure, saying they reduced Nvidia's China market share from 95% to 50% while accelerating China's domestic chip development. During a visit to Taiwan with Lex Fridman, Huang separately argued that broad technology sales generate wealth that funds R&D, sustaining US technological advantage, and framed the policy choice as optimizing to block rivals versus optimizing for a compounding innovation flywheel. China's Commerce Ministry called US chip policy 'overreaching' and 'bullying' and demanded the White House correct its mistakes. The White House scrapped the Biden administration's tiered AI Diffusion Rule and promised a replacement.