
Tech • AI • Robotics • Game
⏱ ≈ 3 min read · 1 h 25 min video — 1 h 22 min saved
Nvidia is facing a leadership and strategy crisis after the ouster of Jensen Huang, exposing a wider fight in the United States over AI, China, and the future of open versus closed models.
Jensen Huang, presented as having stepped down after revelations about multiple affairs, has become the center of a broader power struggle. Colette Kress, Nvidia’s chief financial officer, has taken over on an interim basis and publicly signaled little enthusiasm for his return. Huang later said he had made serious mistakes and would be honored to come back, suggesting the battle is far from over.
The departure has rattled parts of Silicon Valley because many of Nvidia’s relationships reportedly relied on Huang’s personal authority as much as on formal contracts. Some investors have urged the board not to close the door on a comeback, arguing that removing a co-founder of such a critical company could damage long-term value. The key question is whether top talent and major partners stay loyal to Nvidia without him.
Several interpretations now frame the affair less as a personal misconduct story than as a clash over US-China technology policy. One camp favors restricting Chinese access and pushing tightly controlled, proprietary AI systems. Another prefers keeping foreign firms dependent on American infrastructure, especially Nvidia chips, even if that means broader circulation of open-weight models.
A central theory emerging in the debate is that Chinese AI companies want access to the best American chips, while the Chinese government has a stronger interest in technological independence. In that view, selling or enabling access to Nvidia hardware could keep Chinese AI players tied to the US stack and slow the rise of a fully sovereign Chinese ecosystem. That makes Nvidia a strategic instrument, not just a semiconductor company.
The dispute reflects wider tensions inside the Trump administration, where competing factions are said to disagree sharply on trade, China and AI governance. One side sees open-weight AI as a national security risk that should be constrained and monetized through licensed American platforms. The other sees openness as a way to spread US standards, software and chips across the world and preserve technological leverage.
Even amid the leadership drama, Nvidia reported quarterly revenue of $116 billion, roughly double from a year earlier. That performance shows the underlying machine is still running, but critics argue those numbers largely reflect work already set in motion under Huang. A stronger balance sheet may calm markets in the short term, but it does not answer the strategic question of who can lead the company through the next phase of the AI race.
The debate also underscored how little control Europe has over the direction of AI infrastructure. If open-weight models are restricted and access to frontier systems becomes more expensive, European companies may have little choice but to depend on American providers. That prospect is feeding renewed interest in local hosting, smaller deployable models and vendors such as OVHcloud and Mistral, though participants acknowledged the gap with the US remains large.
At the same time, the United States has been hit by a string of serious digital disruptions, including a nationwide grounding order from the FAA, service issues affecting Outlook, Apple Pay, banking systems and other platforms. Some observers suspect coordinated cyber activity, while others point to aging infrastructure and rushed patching of vulnerabilities exposed by AI-assisted security research. Either explanation suggests growing fragility in core digital systems.
A recurring argument is that modern AI systems are uncovering weaknesses in software and infrastructure at a pace older systems were never designed to handle. That creates immediate instability, but it could also force a large-scale rebuild of critical technology stacks. For cloud, cybersecurity and enterprise vendors, this may open a major new market as companies shift from paying mainly for tokens to paying for deployment, integration and secure local operation.
The fight around Nvidia now looks like more than a boardroom scandal. It has become a test of how the United States intends to wield AI, chips and openness as instruments of power.
Ask a question