The disclosure that Jensen Huang, the CEO of Nvidia, engaged in an affair with a former executive of a company based in Singapore has sparked worries about China’s potential to obtain advanced microchips that are currently restricted from entering the country.
Huang has acknowledged his affair with Alice Huang, a Chinese entrepreneur who eventually became the majority shareholder of Megaspeed International, a cloud-computing firm situated in Singapore.
According to a report from The New York Times last month, U.S. officials have determined that she was acting as an agent for Chinese intelligence, sharing sensitive information with Beijing while cultivating a close bond with the Nvidia CEO — claims that the Foreign Ministry of China has dismissed as “groundless.”
Additionally, Megaspeed has come under investigation by U.S. authorities due to suspicions that it facilitated Chinese companies in acquiring advanced Nvidia chips, which are subject to export limitations imposed by Washington.
This scandal has also highlighted questions regarding the effectiveness of U.S. restrictions on these high-end chips destined for China.
Let’s examine the reasons behind China’s interest in these elite chips, the rationale for the U.S. export ban, and the ways in which China continues to procure them.
Why does China want Nvidia’s chips?
Nvidia is renowned for creating the most powerful AI chips globally, which analysts regard as the benchmark in the industry.
“Their chips are leading the way in the AI revolution,” stated Ryan Ellis, an associate professor at Northeastern University in Boston, whose research encompasses communication law, policy, infrastructure politics, and cybersecurity.
According to Dennis Wilder, a former senior U.S. intelligence officer and National Security Council director for China, artificial intelligence demands substantial graphics processing (GPU) power.
High-performance chips from Nvidia offer a cost-effective approach to advancing AI without requiring the establishment of additional data centers to expand GPU capacity, Wilder explained.
These chips are also essential for the most sophisticated AI models, as noted by Ellis, indicating that limited access to them hinders progress.
“There is an immense demand for these chips within China,” he remarked. “A flourishing black market exists for the most advanced video chips in the country.”
Why did the U.S. prohibit sales of these chips to China?
In October 2022, the Biden administration’s U.S. Department of Commerce enacted export controls on advanced computing chips destined for China. This measure means that Nvidia’s top-tier chips, Blackwell Ultra and Vera Rubin — both GPUs designed for AI applications — are barred from being sold to the nation.
The U.S. government cites two primary reasons for these sales restrictions: to maintain the country’s leadership in AI advancement and to safeguard national security.
Whoever leads the AI race — and is capable of training the most effective models — will hold a significant advantage in this forthcoming era of AI innovation, according to Chris Mattmann, the first chief data and AI officer at UCLA.
While the U.S. currently leads China, that gap is narrowing, according to Mattmann. Any factor that could impact this dynamic, such as access to cutting-edge chips, could help further reduce that gap.
The U.S. government is also focused on preserving national security, as there are concerns that granting China access to these chips might enhance its military capabilities, potentially leading to the creation of systems that could one day jeopardize U.S. interests.
“Export controls always aim to foster market growth while simultaneously shaping that market to ensure it does not compromise national security,” remarked Ellis.
However, former U.S. President Donald Trump has relaxed some of the restrictions, now permitting the sale of certain Nvidia chips that were previously barred — typically those used in consumer electronics — to China under specific conditions.
How is China acquiring Nvidia chips?
Shortly after the implementation of restrictions on advanced chips to China, numerous vendors established an illicit smuggling and underground network to funnel these chips into the country for sale to local businesses, as reported by Germany’s Deutsche Welle (DW).
The U.S. has attempted to crack down on these vendors where feasible. In December 2025, the Department of Justice announced the arrest of two Chinese individuals for allegedly smuggling Nvidia chips into China.

A screengrab from CTI video footage shows a woman identified by Reuters as Alice Huang, after comparison with a photograph obtained separately, attending a dinner with Nvidia CEO Jensen Huang at a restaurant in Taipei, Taiwan, on May 22, 2025. (Reuters)
However, there is also a legal loophole that can be exploited to transfer chips to China.
Although Nvidia cannot directly sell its most advanced chips to China or companies based there, it can sell them to other nations, and third-party data centers can subsequently rent access to those facilities to companies with ties to China.
For instance, Nvidia could sell its chips to a Malaysian firm, which could then lease that computing power to a Chinese enterprise, Ellis explained, and “that would be permissible.”
While some U.S. legislators believe this loophole should be eliminated, others argue that it represents a reasonable solution, according to Ellis.
However, Ellis pointed out that the ongoing scandal implies that these chips are frequently sold to other nations, only to be redirected straight into China afterward.
Are the export controls truly effective?
Ellis asserts that U.S. export controls are “never going to be completely foolproof.”
Yet, he believes there could be more random inspections and stricter oversight to ensure that the buyers of these chips remain the ultimate users. “The enforcement is lacking,” he noted.
The Bureau of Industry and Security, the enforcement branch of the U.S. Commerce Department, is not adequately staffed to enforce the restrictions, as Wilder concurred.
For example, Reuters reported that last month, China’s Northwestern Polytechnical University procured Nvidia H200 chips, according to documentation. Other Chinese universities with close affiliations to the defense sector or military research have also acquired Nvidia chips that are prohibited from export.
These export controls have also prompted discussions about whether they are effectively slowing down Chinese technological progress.
Since the U.S. excluded Huawei — and later China — from acquiring some of the leading computer chips and manufacturing equipment in 2019, Chinese semiconductor manufacturers have raced to achieve self-sufficiency, developing their own chips and expertise.
Jensen Huang himself has argued for more relaxed export controls on China, asserting that continued U.S. sales would maintain American technological dominance and ensure Chinese customers remain dependent on U.S. technology.
“We must ensure, firstly, that we have national security and protect our nation, but at the same time, we should compete and advance our technology sector and maximize our exports,” the Nvidia CEO stated to The Associated Press in June.


