Nvidia's $124 Billion Web Meets Regulatory Shock After Huang Exit

Nvidia's equity investments and commitments totaled $124 billion as of the end of July, and the man who personally signed off on much of that network — CEO Jensen Huang — has now left the company under a cloud. That's the number practitioners should hold onto as two shocks land on the AI industry at once: Huang's disgraced exit and a Democratic landslide in the U.S. midterm elections, won by a party that campaigned on data center regulation.

The immediate worry is funding. Nvidia stopped being just a GPU vendor years ago; it became the money pipeline of the AI ecosystem.

• More than 120 venture investments executed this year alone, per the Financial Times
• Over $60 billion (about 80.13 trillion won) poured into 200-plus startups, including OpenAI, Anthropic and Elon Musk's xAI
• $36 billion committed to future purchases of computing capacity from AI cloud providers
• $20 billion sunk into data center leases

The FT noted that a significant number of these deals were decided by Huang personally and relied on his network. The structure drew heavy criticism as "circular financing": companies Nvidia invested in turned around and bought Nvidia GPUs, while Nvidia helped them finance data centers and power.

What does the double shock change?

Two risks now stack. The Democratic sweep raises data center permitting hurdles, power costs and regulatory burdens. The Huang affair adds funding risk for AI infrastructure, since he designed and managed much of the investment chain himself.

Could he come back, Steve Jobs-style? The obstacles look considerable. Huang is the subject of investigations by the U.S. Congress and intelligence authorities over alleged ties to Chinese espionage and suspicions about the possible diversion of AI chips to China — with an internal affair scandal layered on top.

But Nvidia's contracts don't evaporate with its CEO. "Nvidia is a $4 trillion company. Not because Jensen Huang is CEO, but because it posted $100 billion in revenue last quarter and grew 100% from a year earlier," said Matt Witheiler, head of late-stage growth in the private investment division at Wellington Management.

Over the medium to long term, some analysts see the episode diversifying the supply chain instead. AMD, Broadcom and Intel are already pushing to break the dominant position of Nvidia's GPUs.

What does it mean for power demand?

The IMF struck a cautious note. Asked by Reuters about sharp swings in AI stocks and the Democrats' pledge to slow data center development, it said it's monitoring closely but it's too early to conclude whether a broad reassessment of AI technology, investment, profitability and productivity will take place. IMF Managing Director Kristalina Georgieva said in August that AI investment is helping the global economy hold up better than feared amid an energy shock triggered by conflict in the Middle East.

If that holds, the impact on AI-driven electricity demand could be limited. Computing demand wouldn't disappear — it would shift. For power producers and grid operators, the real variable is how much computing the industry consumes overall, not who supplies the GPUs.

Who carries the acceleration argument now?

Huang was the standard-bearer for speed. When slowdown calls gained traction, he argued no new laws or regulations were needed, that companies should run as fast as they can, and that choosing between innovation, speed and safe products is a false choice. On existential risk, he put the probability of humanity's extinction within 10 years at 0% and called doomsday narratives irresponsible.

Set aside whether he was right. One thing is clear: the loudest voice arguing that AI must move faster is gone. The industry ran on the conviction that it needs more computing, more data centers and more power — and Huang represented that conviction most forcefully. Whether a second Jensen Huang emerges is now an open question.