The US Department of Commerce issued new guidance on May 31, 2026 that closes the overseas subsidiary loophole for advanced AI chip exports, marking the most significant expansion of US semiconductor restrictions since the original China export controls began in 2022. The guidance effectively ends NVIDIA’s ability to sell its most advanced chips to Chinese firms through third-country subsidiaries.
The Loophole That Was
The original export controls, implemented in 2022 and strengthened multiple times since, targeted companies headquartered in China. However, they did not explicitly address companies headquartered elsewhere but with Chinese ownership or control. This gap allowed Chinese firms to access advanced NVIDIA chips—including the powerful Blackwell family—through subsidiaries in countries like Malaysia, Singapore, and the United Arab Emirates.
According to reports, hundreds of thousands of NVIDIA Blackwell and AMD MI350x chips flowed to Chinese-controlled entities through this route before the loophole was closed.
The New Rules
The May 31 guidance extends export licensing requirements to cover any entity whose ultimate parent is headquartered or majority-owned in China, regardless of where that entity is physically located. The rules explicitly target advanced processors including NVIDIA’s Blackwell chips and AMD’s MI350x series.
The effective date was May 31, 2026, with enforcement applying to both new sales and existing contracts under review.
The Market Impact
NVIDIA’s share of China’s AI accelerator market has fallen from 95% in 2022 to effectively zero as of June 2026. The company previously derived billions in revenue from Chinese data centers but has been shut out of the market entirely.
Analysts estimate that the total revenue opportunity lost exceeds $17 billion annually. NVIDIA has shifted its China strategy toward lower-end chips that don’t require export licenses, but these represent a fraction of the AI accelerator market.
China’s Response
Chinese technology firms are accelerating domestic chip development in response. Huawei’s Ascend series and other domestic AI accelerators are receiving increased investment as the country seeks self-sufficiency in AI infrastructure.
The US move is also prompting calls in Europe and Asia for greater chip independence, as the export controls raise concerns about supply chain reliability for AI computing infrastructure.
What This Means
The closure of the overseas subsidiary loophole marks the end of an era where Western AI chips flowed to China through any channel. The AI race is increasingly becoming a bifurcated ecosystem—with China developing its own stack independent of Western technology.
For enterprises building AI infrastructure outside China, the supply constraints may ease as NVIDIA focuses on serving the unconstrained market. But the broader implications for global AI development are significant: the technology decoupling that began with export controls is now nearly complete.