US allows Nvidia H200 chips to be exported to China

US allows Nvidia H200 chips to be exported to China is profiled by BTW Media because published evidence links it to internet infrastructure, governance, operational dependencies, or market visibility.

• The US administration, under Donald J. Trump, has green-lit the sale of H200 chips to China, subject to export licences and national-security safeguards.
• The decision has raised concerns in Washington, with lawmakers preparing legislation to restrict such chip exports to adversary countries — the new policy may face legal and political challenges.

What happened: US reverses AI-chip ban, approves H200 exports to China

In a significant reversal of previous policy, the US administration announced on 8 December 2025 that it will permit Nvidia to export its H200 artificial-intelligence chips to approved customers in China. The chips, which underpin cutting-edge AI workloads, had been barred from export under earlier US export controls. The newly issued policy allows shipment under strict licensing and oversight, with the US reportedly to take a 25% share of chip-sale proceeds.

The decision follows diplomatic developments between Washington and Beijing, including a recent truce in their broader trade and tech dispute. While the export clearance covers the H200 chip, it does not extend to Nvidia’s newest GPU architectures like Blackwell or Rubin — which remain restricted.

Nvidia’s shares rose modestly on the announcement, reflecting investor optimism that access to China’s vast AI market could restore some of the company’s lost revenue. Company executives have welcomed the move but cautioned that final export licenses and careful compliance procedures must be followed.

Also Read: China bars ByteDance from using Nvidia chips in new data centres
Also Read: Microsoft, NVIDIA and Anthropic Forge Landmark AI Partnership

Why it’s important

This policy shift marks a major turning point in the global AI arms race. Granting China access to powerful H200 chips could accelerate Chinese firms’ capabilities in large-scale AI model training, narrowing the hardware gap between China and the US. For Nvidia, renewed access unlocks a key market — but only under tight US controls.

The decision has already provoked concern among US lawmakers: a bipartisan bill introduced shortly after the announcement seeks to reinstate export restrictions for high-end chips to rival states, citing national-security risks.

Beyond geopolitics, the move could reshape global AI infrastructure. Chinese data centres and research labs may now ramp up investment in AI compute power, potentially shifting more AI workloads eastward. Meanwhile, the shift raises urgent questions around supply-chain transparency, end-use controls and the ethics of global AI hardware distribution.