- Leadership Change: Jensen Huang resigned as Nvidia CEO, with finance chief Colette Kress taking interim charge of the AI-chip maker.
- Congressional Demand: House committee leaders sought a briefing from Huang on Nvidia’s espionage protections and discussed subpoenas if he did not respond.
- Chip Exports: Senators want congressional review of advanced-chip sales to China included in annual defense legislation.
- Overseas Computing: Some Chinese AI firms are reportedly seeking overseas Nvidia computing capacity amid fears of tighter U.S. restrictions.
Jensen Huang resigned as Nvidia’s chief executive on November 12 amid a congressional inquiry into the company’s safeguards against alleged Chinese espionage. Finance chief Colette Kress took interim control of the AI-chip maker, whose advanced processors are widely used to train and run artificial-intelligence models.
Huang also resigned as president and a director, effective immediately. Nvidia’s transition announcement gave no reason for his departure. Kress will retain her finance responsibilities while the board searches for a permanent successor, considering both internal and external candidates. Harvey Jones became the company’s first board chair.
The departure came a day after Republican Senator Kevin Cramer and Representatives Suhas Subramanyam, a Democrat, and Don Bacon, a Republican, called for Huang to resign in CNBC interviews. Cramer argued that Huang’s affair with an alleged Chinese spy could expose him to extortion or blackmail, creating a national-security risk.
Congress Seeks Answers About Espionage Risks
Huang acknowledged an affair on November 8. The woman, Alice Huang, is a former executive at Singapore-based data-center company Megaspeed, which advertises a partnership with Nvidia. The two Huangs are not related. The New York Times reported that U.S. intelligence agencies determined she was passing information to Beijing.
China has denied the spying allegations. Nvidia’s board had already opened an inquiry into the relationship and said it had no reason to believe national security or company secrets had been compromised. President Donald Trump said Huang had told him he did not know she was an alleged spy.
House Foreign Affairs chair Brian Mast, China committee chair John Moolenaar and Intelligence chair Rick Crawford requested a joint closed-door briefing on Nvidia’s security protocols and Huang’s interactions with Alice Huang. Their November 4 letter proposed a November 18 appearance and asked him to accept or offer another date by November 11.
On November 11, Mast’s spokesperson said he was open to issuing a subpoena if Huang did not respond. Moolenaar also described subpoenas as a likely response if Huang neither agreed to meet nor proposed another date. Huang’s response after that deadline and any subsequent subpoena were unconfirmed on November 12.
Megaspeed has denied illegally diverting Nvidia chips to China. On November 12, it said it had cooperated with the Commerce Department’s Bureau of Industry and Security, which enforces technology export controls, and would cooperate with other U.S. authorities. The company also said it had hired a U.S. law firm to conduct a chip census and verify its export controls.
How the Controversy Could Shift Chip Policy
Huang has long argued that China already has computing power for AI development and that broader adoption of American technology would help preserve U.S. dominance. Restriction advocates emphasize denying China advanced hardware for national security.
That disagreement predates the personal allegations. In September 2025, Senator Jim Banks proposed giving American companies priority access to AI hardware before exports to Chinese businesses, a measure Nvidia opposed.
Rui Ma, a technology consultant and founder of Tech Buzz China, told Al Jazeera that weaker influence for Huang could strengthen restriction advocates regardless of the inquiry’s findings. Erich Grunewald of the Institute for AI Policy and Strategy separately said the controversy could focus attention on loopholes and existing bills.
Senators Elizabeth Warren, a Democrat, and Banks, a Republican, have urged colleagues to include the AI Overwatch Act in the annual National Defense Authorization Act. CNBC reported on November 11 that the defense bill remained unpassed and identified AI Overwatch, Chip Security and MATCH as potential measures to tighten technology controls.
The introduced House version of AI OVERWATCH, dated December 18, 2025, would give Congress a role in individual license decisions for covered data-center chips sent to entities located or headquartered in China or other countries of concern. The same review would apply if the recipient’s ultimate parent company, at the top of its corporate group, is headquartered there. Commerce would submit applications and security certifications at least 30 days before approval. Congress could prohibit a transaction by enacting a joint resolution during that period. The certifications would address military or intelligence uses, adequate U.S. supply and America’s computing advantage.
The introduced Chip Security Act, dated May 15, 2025, addresses what happens to controlled chips after shipment. It would direct Commerce, within 180 days of enactment, to require feasible and appropriate location-verification mechanisms before export or transfer abroad. Authorized exporters would have to report credible information about diversion to another location or user, or tampering with security mechanisms.
The introduced MATCH Act, dated April 2, 2026, targets equipment and servicing used to manufacture semiconductors. It seeks aligned restrictions from allied supplier countries, with a 150-day deadline after enactment for U.S. rules and an assessment of allied alignment. If allies failed to align, additional U.S. controls would extend to specified foreign equipment and servicing, subject to a conditional extension of up to 90 days.
Gabriel Wildau, a China political-risk analyst at Teneo, urged skepticism because of anonymous leaks and possible motives to discredit Huang and Nvidia. He warned that Beijing could regard U.S. restrictions on remote computing as violating the countries’ trade truce. Chris Clague of the International Institute for Strategic Studies argued that the controversy weakened a legitimate case for technology transfers, but compromise should continue.
Chinese Firms Seek Computing Capacity Abroad
Reuters reported on November 12, citing three people familiar with the matter, that companies were trying to rent Nvidia-powered capacity at overseas data centers because they feared tighter U.S. restrictions on remote access.
The processors stay in overseas data centers while Chinese customers rent the computing power needed to train and run AI models. One major Chinese cloud provider had discussed locking in capacity internally, while some companies had contacted operators about reservations. Smaller AI labs were considering similar steps, including approaching providers in Southeast Asia. Reuters’s sources withheld both their own identities and that of the major cloud provider.
Traders and intermediaries were also stockpiling restricted Nvidia processors, according to two of Reuters’s sources, because buyers feared tougher enforcement against smuggling AI chips and servers into China.
