After Jensen Huang no longer serves as the CEO of NVIDIA, the next leader of the company will face challenges that are vastly different from those that helped the company become one of the most valuable in the world back then.
In addition to maintaining the demand for the AI chip, the new leadership also needs to deal with the increasingly stringent export restrictions from the United States, the investigation into NVIDIA's business in China, and the scrutiny of its financing arrangements.
This is the view of NVIDIA's investor, REX Financial, the Managing Director Bill Birmingham. REX Financial is the parent company of REX Shares, which manages assets worth $8.5 billion.
"The bigger issue in the future is not who will sit in the position of CEO, but rather the issue of China's access to chips," he told Fortune via email.
For many years, U.S. restrictions have been aimed at limiting China's purchase of NVIDIA's most powerful processors. However, an extramarital affair between Jensen Huang and a senior executive from the Singaporean technology company Megaspeed has raised concerns among outsiders about Chinese companies acquiring such restricted hardware.
The Trump administration is now attempting to plug these loopholes. Last Sunday, the federal government announced a rule to close the loophole that allows remote access to advanced AI chips and servers through overseas data centers. On Friday, U.S. President Donald Trump stated that he hopes to prevent American companies from building data centers overseas.
"Whichever company takes over NVIDIA will inherit a more limited and strictly regulated international business, as well as the subsequent complete reorganization of the AI balance of power," said Birmingham.
NVIDIA's latest financial report shows that despite this series of news events, its business momentum remains strong. NVIDIA announced revenue of $115.8 billion for the third quarter of fiscal year 2026, a 103% increase from the same period last year. Of this, revenue from its data center business was $107.6 billion, accounting for approximately 93% of the total revenue.
However, Birmingham believes that the importance of financial reports themselves is declining, and the market should pay more attention to management's statements regarding future demand, pricing capabilities, and the sustainability of the AI investment cycle.
A company is facing a thorough investigation.
The California Attorney General's Office has issued a subpoena to NVIDIA to investigate whether the company or Jensen Huang assisted China in obtaining advanced chips or confidential information, or whether they misled investors regarding the disclosure of risks.
According to the New York Post, the U.S. Securities and Exchange Commission (SEC) is also investigating whether the company misled investors in its public disclosures. As officials question whether NVIDIA was aware that its chips might be re-exported to China and whether its disclosures adequately explained the related risks, this civil investigation is progressing.
Trump also stated that the Federal Bureau of Investigation is investigating the relationship between Jensen Huang and Alice Huang, as well as any connections with Chinese intelligence agencies. The Beijing authorities have denied these allegations.
Congress will also join the review. Jensen Huang and NVIDIA's Chief Legal Officer Tim Teter have agreed to appear before three committees of the House of Representatives to discuss the role the company may play in allowing Chinese entities to obtain advanced American chips.
After the House Foreign Relations Committee disclosed on Thursday that the U.S. Department of Commerce, between December 2025 and May 2026, approved the export of 879,868 advanced semiconductors to Chinese companies through 21 licensing applications, lawmakers will also examine the role played by the Department of Commerce in this process.
AI Infrastructure issues also exist
Birmingham indicates that export restrictions are just one of the challenges that NVIDIA needs to address. The company also has to deal with the financial structure that supports the unprecedented expansion of the infrastructure behind AI.
He is concerned that the demand for NVIDIA processors is closely related to whether hyperscale cloud service providers and smaller cloud providers have the capability to finance substantial investments in data centers. This has created a fragile relationship between NVIDIA's hardware sales and the financial health of its customers.
"Given that the AI ecosystem relies heavily on supplier financing and bond market support, the risk of a collapse similar to that in the telecommunications industry is quite high," he warned.
Birmingham It was announced in August that NVIDIA planned to collaborate with large financial institutions such as BlackRock and Goldman Sachs, as well as KKR, to mobilize over $500 billion in third-party capital for the financing of AI infrastructure, a move that has raised concerns. He stated that these commitments were "clearly" made "solely by Jensen Huang," which means investors need to figure out whether the company intends to proceed with these arrangements.
In the latest quarterly report, NVIDIA disclosed that these financing arrangements are still in the preliminary stage and may not necessarily lead to final agreements. The company also stated that it can provide limited residual value support for specific projects at its own discretion.
Birmingham says that, unlike mature hyperscale cloud service providers, new cloud service providers have less flexibility in dealing with the continuously rising costs of hardware and financing.
"They need nearly perfect utilization rates to remain profitable," he added, "Generally speaking, they can only pass on the higher costs to new contracts, and that's only possible when computing power remains scarce."
This means that if demand weakens, new cloud service providers may find themselves with expensive computing infrastructure but unable to generate sufficient revenue to repay their debts.
"Investors are assessing whether the supply cycle of AI is overheating, and whether there will be a turning point in the demand for computing power that forces a slowdown in data center construction, leading to a self-reinforcing decline throughout the system," he explained.












