Jensen Huang built the machine that powers the age of artificial intelligence. On Thursday he was forced to step away from it. NVIDIA’s board announced that the company’s co-founder, president and chief executive had resigned with immediate effect, ending a 33-year reign that turned a graphics-chip startup into the indispensable arsenal of the AI era. Colette Kress, the long-serving chief financial officer, takes over as interim chief executive while remaining CFO. Harvey C. Jones, a board member since the company’s founding year, becomes its first-ever chairman. The official language was polite. The reality is not.

Huang did not leave because the product roadmap stalled or because customers lost faith. He left amid a cascade of revelations about extramarital affairs—one with a woman American intelligence officials regard as a suspected asset of Chinese intelligence, another with a senior NVIDIA executive who reported directly to him. The board had opened an investigation. Lawmakers had begun to mutter. The stock had already begun to bleed. By the time the formal announcement landed, the outcome was inevitable. Corporate America still prefers the word “resigns.” In practice, this was an ejection.

The company’s statement thanked Huang for “immeasurable contributions” and declared NVIDIA one of the most consequential technology firms on earth. Both claims are true. Under Huang the company became the toll bridge of the AI boom: its GPUs and CUDA software platform sit at the center of almost every serious training run and inference cluster. Hyperscalers, startups, governments and militaries built their strategies around NVIDIA’s cadence. Huang himself became the public face of the industry—leather jacket, theatrical product launches, easy access to presidents and prime ministers. That concentration of power was always a risk. On Thursday the risk crystallized.Kress is the safe pair of hands. She joined in 2013 after long stints at Microsoft and Cisco, and has been the steady voice on earnings calls through the company’s most vertiginous growth. Jones, a veteran of Synopsys and Silicon Valley boardrooms, supplies institutional ballast.

The board insists operations, customer commitments and the product roadmap will continue without interruption. Markets were less convinced. Shares dropped more than 10 percent in premarket trading, extending a multi-day slide that has already erased hundreds of billions in value. Investors understand the difference between a founder-led company and a company that has just lost its founder.Analysts reached for the usual vocabulary of key-man risk and tried to sound measured. Gil Luria of DA Davidson called Kress “well respected” and capable of guiding the firm through a difficult period, while noting that Huang had concentrated decision-making to an unusual degree. Dan Ives had earlier described Huang as the “central nervous system” of the American technology ecosystem and warned that his abrupt departure could inflict serious damage.

Stacy Rasgon of Bernstein put it more bluntly: “Jensen is Nvidia.” Josh Gilbert of eToro suggested Kress could steady the ship in the short term, but argued that a permanent chief executive would likely need deeper engineering credibility—names such as CUDA architect Ian Buck have already begun circulating. The permanent search is open to internal and external candidates. That process will be watched with unusual intensity.On X the reaction was faster and less polite. Users tracked the premarket plunge, mocked the sudden end of the leather-jacket era, and zeroed in on the national-security dimension.

Some framed the episode as a cautionary tale about elite capture and Beijing’s patient influence operations. Others focused on the practical question: who now decides the next generation of AI hardware while America and China race for technological primacy? A few noted the irony that a company so central to Western AI leadership had allowed its most important executive to become entangled in relationships that invited exactly this kind of scrutiny. The corporate communications team’s careful language did not travel well in that environment.

The deeper story is not merely personal failure. It is about power, concentration, and the uncomfortable intersection of private conduct and public technology strategy. NVIDIA sits at the heart of the U.S.-China technology competition. Export controls, chip design leadership, and the ability to supply the world’s most advanced accelerators are strategic assets, not just commercial ones. When the man who personified that leadership becomes the subject of intelligence concerns and boardroom investigations, the consequences extend beyond one company’s share price.

Washington has spent years trying to restrict advanced chips from reaching Beijing’s military and surveillance apparatus. The optics of the chief executive of the dominant supplier becoming entangled with a figure linked—by American officials—to Chinese intelligence are, to put it mildly, suboptimal.Huang’s defenders will argue that private failings should not erase a historic industrial achievement. They have a point. Few executives have created more economic value or done more to advance a foundational technology. The company’s engineering culture, developer ecosystem and product pipeline are not going to vanish overnight. Yet modern corporate governance, especially at firms of systemic importance, has grown less tolerant of the old distinction between the boardroom and the bedroom when national security and market confidence are in play.

The second alleged relationship—with a subordinate—made the board’s position even more difficult. Once that detail surfaced, the politics inside the company and in Washington hardened.What comes next will test whether NVIDIA can institutionalize the advantages Huang accumulated through force of personality and strategic intuition. Kress and the existing leadership team must demonstrate that the machine still runs without its most visible operator. Customers will demand reassurance on delivery schedules and architectural direction. Competitors—AMD, custom silicon efforts at the hyperscalers, and China’s domestic champions—will probe for any hesitation. Policymakers will watch the firm’s posture on export compliance and engagement with the administration more closely than before.

The AI boom has always rested on a narrow set of industrial and intellectual foundations. NVIDIA was the most important of them. For three decades the company and its founder moved in lockstep. That era is over. The leather jacket has been hung up. The question now is whether the institution can match the vision of the man who built it—while operating under a harsher spotlight than Huang ever faced. Markets, rivals and governments will render their verdict in the weeks and months ahead. The board has chosen continuity and process.

History will decide whether that was enough.