Trump Administration Tightens AI Chip Controls Amid Nvidia CEO Scandal
The Trump administration has significantly expanded U.S. export controls on artificial intelligence technology, targeting both the physical shipment of advanced chips and the remote computing services used by Chinese companies. The policy shift comes shortly after Nvidia CEO Jensen Huang resigned following revelations of an affair with a woman linked to Chinese intelligence.
The Trump administration has implemented new measures to restrict Chinese access to advanced artificial intelligence chips, targeting not only physical hardware but also the remote computing services that have allowed Beijing to bypass previous sanctions. The announcement on Monday introduced licensing requirements for overseas data centers providing significant computing power to China and imposed stricter rules on chip shipments to Southeast Asian nations where officials suspect smuggling activities.
The timing of these policy changes has drawn intense scrutiny, coinciding with the resignation of Nvidia CEO Jensen Huang last week. Huang stepped down after admitting to an extramarital affair with Alice Huang, a Chinese businesswoman who previously worked for Megaspeed, a Singapore-based data center company suspected of facilitating chip smuggling. While the White House stated that the new rules are designed to prevent the smuggling of advanced U.S. silicon and ensure that innovation benefits American companies first, observers note the proximity of the policy shift to the personal scandal involving the industry’s most prominent figure.
The current administration’s actions represent a significant escalation in the United States’ long-standing effort to curb China’s technological advancement. The strategy began in the fall of 2022 under the Biden administration, which introduced export controls aimed at choking off China’s access to the high-performance chips required to train large AI models and the equipment needed to manufacture them. At the time, the approach was considered radical because it sought to leverage U.S. dominance in the global semiconductor industry to slow the world’s second-largest economy, moving beyond traditional export controls focused on military programs or human rights abuses.
The core logic of the strategy relied on the fact that training advanced AI models requires immense computing power, a market dominated by American firms like Nvidia. By restricting access to these chips, Washington hoped to preserve its technological lead. However, four years later, the effectiveness of these controls has been complicated by enforcement challenges and adaptive behaviors by Chinese entities.
A primary obstacle has been the Bureau of Industry and Security (BIS), a relatively small agency within the Commerce Department tasked with enforcing the rules. With a budget of just $235 million in 2026, the BIS faces significant resource constraints compared to the revenues generated by major chipmakers. As the Biden administration introduced successive rounds of complex restrictions, Chinese companies and intermediaries developed increasingly elaborate methods to circumvent them.
Smuggling has become a substantial issue. Megaspeed has emerged as a notable example of this problem. According to Bloomberg, approximately 50,000 of the estimated 136,000 AI chips imported by Megaspeed into Malaysia and Indonesia in 2024 and 2025 were not accounted for in Nvidia’s initial inventory checks. While Nvidia later stated it had located some of these chips, researchers at the Institute for Progress estimate that China may have accumulated smuggled computing power equivalent to roughly one million high-end Nvidia H100 chips by the end of 2026.
Beyond physical smuggling, Chinese companies utilized a method that did not violate existing export controls: renting computing power from data centers overseas. This practice, known as remote access, allowed firms like Alibaba and ByteDance to use powerful Nvidia chips without physically importing them into China. Policymakers were aware of this loophole but initially tolerated it, arguing that forcing Chinese firms to rely on American technology overseas might keep them dependent on U.S. systems rather than pushing them to develop domestic alternatives. Additionally, restricting remote access was seen as difficult to enforce and potentially harmful to U.S. technology companies.
The new rules announced by the Trump administration mark a shift in this approach. By requiring special licenses for overseas data centers providing large amounts of computing power to China, the White House is attempting to regulate how Chinese entities use American technology regardless of its physical location. This expansion of control follows previous efforts that focused solely on preventing the acquisition of specific chips.
The impact of these restrictions remains a subject of debate. Evidence suggests that export controls have made it harder for Chinese AI companies to acquire computing power comparable to their American competitors. However, assessing the precise effect has proven difficult. The restrictions have also imposed costs on U.S. firms. A study by the Federal Reserve Bank of New York found that earlier export restrictions wiped approximately $130 billion from the combined market value of affected American suppliers, while also hurting profitability and employment.
Furthermore, the pressure has accelerated China’s efforts to develop independent capabilities. Huawei has developed methods to link powerful chips into larger megachip systems, and Chinese AI lab DeepSeek has demonstrated that efficient training techniques can allow developers to build competitive models with fewer resources.
Critics argue that the Trump administration may be overextending its authority by attempting to control not only which chips Chinese companies can buy but also how they use American technology globally. The decision to tighten controls in the wake of Jensen Huang’s personal scandal has raised questions about the motivations behind the policy shift and whether such broad regulatory power is appropriate given the circumstances.
As the U.S. continues to navigate this complex landscape, the balance between national security interests and economic consequences remains a central challenge. The new rules will require overseas data centers to navigate additional red tape, potentially altering the global flow of computing resources and further straining relations with countries like Singapore and Malaysia that serve as hubs for international tech infrastructure.
S.Sadoun--al-Hayat