Trump Allows Nvidia AI Chip Sales to China with 25% Government Cut

A Strategic Reversal in Tech Policy

The Trump administration has officially approved Nvidia's export of its powerful H200 artificial intelligence chips to China, marking a dramatic shift from the previous administration's restrictive approach. President Donald Trump on Wednesday said that his administration will approve China sales of Nvidia's H200 chip for artificial intelligence, but the U.S. government will take 25% of sales , creating an unprecedented revenue-sharing arrangement between a private company and the federal government.

This decision reverses years of export restrictions that had blocked advanced AI semiconductors from reaching Chinese markets. The move marked a significant shift in US export policy for advanced AI chips, which Joe Biden's administration had heavily restricted over national security concerns about Chinese military applications . The policy change reflects Trump's broader approach to trade negotiations, where economic leverage takes precedence over outright prohibition.

Strict Conditions and Safeguards

The approval comes with stringent requirements designed to protect American interests and prevent military applications. Exports will be allowed to "approved customers" with security safeguards and China won't be allowed to use the chips for military purposes and is not allowed to import more than 50% of the chips sold to US customers . Additionally, the H200 chips must undergo a third-party review before being exported to China .

The Commerce Department has established clear performance thresholds for eligible chips, ensuring that only processors operating below specific technical limits qualify for export. Export approvals will be denied if the chips are intended for military, military-intelligence, nuclear, missile, or chemical and biological weapons applications . These safeguards aim to balance commercial interests with national security concerns.

Market Dynamics and Competition

The H200 represents a strategic middle ground in Nvidia's product lineup. Unlike Nvidia's previous China-targeted chip, the H20, the H200 is a version of the company's Hopper generation that is also sold in the U.S. and in other markets. The H200 was not specifically designed and slowed down for export . However, H200s are roughly 18 months behind the US company's most state-of-the-art offerings, which will still be off-limits to China .

Chinese demand for these chips remains substantial despite domestic alternatives. Last week, Nvidia CEO Jensen Huang told reporters that the company is seeing "very high" interest in its H200 chip from China customers, and that the company had started producing the chip again . Yet uncertainty persists about actual market uptake, as Beijing has reportedly been encouraging tech companies to use homegrown chips .

Long-term Implications for AI Leadership

This policy shift reflects deeper questions about America's strategy in the global AI race. Nvidia chief executive Jensen Huang has advocated for the company to be allowed to sell some of its more advanced chips in China, arguing the importance of AI systems around the world being built on US technology . The approach suggests that maintaining technological influence through controlled access may prove more effective than complete isolation.

The revenue-sharing model could establish a template for future tech trade negotiations. "It will be interesting to see if this tariff model expands to other sectors" , notes one industry analyst. As China continues developing domestic chip capabilities and American companies seek global market access, this arrangement may represent a new paradigm for managing technological competition between superpowers while preserving strategic advantages.