Five days after Jensen Huang resigned, Nvidia posted a record quarter. Then its interim chief executive said something every founder with Nvidia on their cap table will want to read twice.
The chipmaker has backed almost a third of AI unicorns since 2023, and the people who track its money now expect its biggest cheques to stop. Washington, meanwhile, has already moved against it.
Jensen Huang’s exit from Nvidia, the company he co-founded in 1993 and led for 33 years, has sent a shock wave far beyond the firm itself. Nvidia’s board accepted his resignation last week, after reports that he had an extramarital affair with a woman US intelligence agencies believe was a Chinese asset. CFO Colette Kress is now interim CEO, and longtime director Harvey Jones has become the board’s first-ever chair.
On the day Huang’s resignation was announced, Nvidia lost about $182 billion in market value, closing down 4.3% after falling as much as 14% intraday. The shares are now about 30% below their October 6 record high.
Nvidia has become one of the AI economy’s biggest backers, committing more than $40 billion to AI equity investments in the first 4 months of 2026 alone, and how it deploys its capital now, in the face of this shake-up, affects everyone: startups, venture firms, regulators and retail investors.
Its first results without Huang, released on 17 November, gave little sign of a slowdown. Revenue for the quarter to 25 October reached $115.8 billion, up 103% year on year, and data centre revenue hit a record $107.6 billion. But on the earnings call, Kress said Nvidia would be “rigorously evaluating future investments, commitments, and guarantees,” and would remain “disciplined in how we commit capital.”
The venture machine
According to PitchBook, Nvidia has backed almost one third of AI unicorns since 2023. The company has also joined a VC deal every third day in 2026. The rounds Nvidia joined are worth a combined $219.5 billion, which accounts for 27.2% of global venture deal value, up from 7.1% just 1 year ago.
Nick Patience, vice president and AI practice lead at The Futurum Group, said that Nvidia’s investment strategy is independent of who is at the helm. “Nvidia’s venture activity has always been about creating demand for its own platform,” he tells Tech Funding News. Every startup it backs is “a future buyer of its systems and another team building on CUDA”, its primary computing platform.
What changes, Patience argues, is appetite for the largest stakes. “Those were Jensen’s deals, often built on personal relationships,” while the interim CEO has to conduct business operations in the midst of an investigation. “Smaller venture cheques will keep flowing,” he expects, “but the nine- and ten-figure commitments are likely on hold until there’s a permanent CEO.”
Kaidi Gao, a senior analyst at PitchBook, arrives at a similar conclusion. In a report shared with the media, Gao sees continuity in Nvidia’s “broad venture strategy”. Although Nvidia’s capital is mainly concentrated in the frontier AI labs, it still spreads investment across every stage. Gao writes that “CFO Colette Kress serving as interim CEO and seasoned venture investor Harvey Jones serving as the board’s first-ever chair” are positive signs this approach will continue.
What it means for Europe
The pause Patience describes would land hardest on Nvidia’s European investments. Its $12.9 billion Hugging Face acquisition came with plans for further investment in Mistral and Black Forest Labs, and in October it pledged $250 million of compute to UK AI safety research and planned stakes in Boltz, Foresight and Hiverge, with terms still to be agreed. Whether those commitments survive Kress’s review is the question European founders will be asking.
Washington moves first
Chris Miller, a professor at the Fletcher School at Tufts University, recently argued in the Financial Times that this Nvidia scandal “will significantly reduce its influence over Washington’s technology policy.” He describes the company’s lobbying as “uniquely confrontational.” Rather than making routine commercial arguments for China sales, Huang attacked the premise of US export controls. This alienated lawmakers in both parties.
Patience says the shake-up has already had an effect. On Monday, the Commerce Department issued a rule to close the “remote-access loophole,” which let Chinese firms use advanced chips housed in data centres abroad. It also added licence requirements for shipments to Singapore, Malaysia, Indonesia, Thailand and Brazil.
“That rule had been in the works for a long time, and this scandal gave it the push it needed. I expect Congress and China hawks in both parties to push for more automatic enforcement with less room for discretion at the top.” Meanwhile, the President remains publicly supportive of Huang and is about to meet Chinese leader Xi Jinping, so “policy and politics may pull in different directions for a while,” says Patience.
Yet, the financial exposure to China looks limited. Cao notes in her report that “[c]ustomers headquartered in China (including Hong Kong) accounted for 7% of Nvidia’s revenue in the six months ended July 26, 2026.” PitchBook found that Nvidia has made no direct investments in companies headquartered in China, Hong Kong or Malaysia since 2020.
Will rivals benefit?
Cao is sceptical that the leadership transition itself accelerates a push away from Nvidia’s chips, towards companies that offer alternatives.
“The largest cloud providers (hyperscalers) and frontier AI labs already have reasons to diversify their chip suppliers,” she tells Tech Funding News. “Nvidia’s chips are expensive, relying on a single supplier is risky, and alternatives give buyers pricing leverage.” Huang’s absence does not change these factors, in principle.
Nvidia’s data centre arm generated $107.6 billion in the quarter ended October 25, about 93% of revenue, and hyperscale revenue alone reached $60.1 billion, according to the company’s latest 10-Q. Cao also cites the fact that “sales to hyperscalers more than doubled,” even though several are designing their own chips.
The real variable, she says, is spending. The question as to whether AI revenue can grow fast enough to justify the expenditures on data centres and chips has been hanging in the air for a while now. “If it does, demand for Nvidia’s AI chips and systems should stay strong.” If it falls short, however, spending could slow and much cheaper alternatives could capture what remains of the market share.
Bubble or concentration?
Although many articles have linked Huang’s resignation to the AI bubble, both Patience and Cao believe these are separate issues. As stated above, the real metric to follow is future revenue when compared to the gargantuan infrastructure costs.
What the episode does expose, he said, is concentration. NVIDIA is about 8% of the S&P 500, so index-fund investors carry “a single-name, single-person exposure most of them didn’t know that much about.” That, he argued, is “more about portfolio management than evidence of a bubble.”
“What this does expose is concentration risk,” says Patience. “One company makes up an unusually large share of the S&P 500 and its value has been tied closely to one person. That means ordinary investors in index funds are carrying a single-name, single-person exposure most of them didn’t know that much about. That’s more about portfolio management than evidence of a bubble, I’d argue.”
Cao offers a reminder that “Like any stock, Nvidia’s price reflects expectations for future earnings, and the CEO transition adds uncertainty to them.” The potential bubble has yet to pop, and it will take more than Haung’s exit to do so. For startups, the more immediate question is whether the world’s most active corporate investor keeps writing cheques while it searches for a leader.