Who would have imagined that Jensen Huang—the man who propelled Nvidia, and indeed the entire United States, to the pinnacle of the global AI industry—would ultimately fall not to commercial competition, but to a sex scandal?

Huang's problem is clearly not his private life.

From Bezos to Musk, sex scandals among American tech titans are hardly uncommon, yet they rarely cost them their business empires.

What Huang has truly lost is Washington's trust. At this point, who he has been involved with—and how many people—no longer matters.

Musk is also facing trouble.

In this midterm election, Musk spent a record $474 million, yet still could not prevent his home base of Texas from turning blue.

As the political landscape shifts, the Department of Government Efficiency (DOGE) led by Musk, SpaceX's government contracts, and the use of Starlink in the Russia-Ukraine war may all face stricter congressional scrutiny.

Behind this, a very important shift in tech investment is actually being revealed:

As the AI industry grows ever larger, the position of these entrepreneurs in American politics has become increasingly delicate. Those political battles—carrying surging public opinion—are becoming the most unpredictable variable shaping the direction of the AI industry.

Compared with visible external threats, invisible internal political risks are often more lethal.

In the face of this risk, the roles of traditional PE, EPS, and moats will ultimately be limited. After all, commercial competition has winners and losers, but political maneuvering can change the rules of the game.

With Huang's downfall, at the very least Nvidia's odds and payoffs need to be reassessed. And this may only be the beginning.

/ 01 / Huang's problem has never been his private life

In the American business environment, almost all U.S.-listed companies strictly prohibit executives from having undisclosed intimate relationships with subordinates. But rules, as always, have traditionally been meant for ordinary people.

If you don't believe it, take a look at Bezos and Musk.

In 2019, Amazon boss Bezos had an affair and eventually went through a divorce. Musk goes without saying: first he was the other man in the Depp couple's marriage, and later he sexually harassed a flight attendant, with the company paying $250,000 on his behalf to settle.

And the result? In the same year as the affair, guess what Bezos's approval rating was in Amazon's board election—nearly 99%.

Musk was even more brazen. In a 2025 official Tesla document, it was written in black and white: shareholders have repeatedly petitioned over the years, hoping the company would retain and motivate our "Emperor Musk."

The board knew the math better than anyone: finding another Musk is far harder than finding another good person.

So Huang's real trouble is not there at all. What he has already lost is Washington's trust.

Many people's first reaction: the Democrats did it, right? After all, the Democrats have been calling for tighter exports of advanced AI chips.

On October 29, The New York Times disclosed U.S. intelligence agencies' suspicions about Alice Huang and her personal relationship with Jensen Huang. On November 3, the U.S. midterm elections ended, and the Democratic Party won control of both the Senate and the House.

But many people did not notice that the three House committee chairs who demanded Jensen Huang go to Congress for a closed-door briefing were all from the Republican Party. Clearly, this time both Republicans and Democrats seemed unprepared to let him off easily.

In fact, the relationship between Jensen Huang and Alice Huang was not some newly discovered secret. At least since 2024, there have been photos of the two appearing together publicly.

Megaspeed, the Singapore company associated with Alice Huang, had long been on the radar of U.S. government investigations.

In February 2025, Singapore police began investigating Megaspeed over suspected false declarations in server procurement and other issues. In June of the same year, the U.S. Department of Commerce also conducted post-shipment verification.

In Nvidia's second quarter of fiscal year 2026, by customer billing address, Singapore contributed 22% of revenue. Just think about it: how could tiny Singapore possibly absorb so many orders? Where the goods ultimately went is no secret in the industry.

Since they knew long ago, why now?

Over the past few years, Jensen Huang has been doing one thing: trying to use his business influence to secure greater operating room for Nvidia in the U.S.-China tech competition.

He repeatedly stressed that restricting U.S. chip exports to China would not stop China from developing AI, but might instead push Chinese companies to break away from the U.S. technology ecosystem and accelerate the development of their own alternatives.

Whether Republican hardliners or Democrats, they had long been gritting their teeth at Jensen Huang.

Before the election, Trump could still shield him. With the end of the 2026 midterm elections and the Democratic Party winning control of both the Senate and the House, the Trump administration's room to maneuver before Congress in the future will also be constrained.

Trump can continue to adjust some export policies through executive power, but he cannot unilaterally stop congressional investigations, hearings, or new restrictive legislation.

But Jensen Huang may have underestimated one thing: when a tech company's products are elevated to the level of national strategic assets, the commercial freedom it can enjoy is no longer entirely determined by the market.

Some things weigh nothing until they are put on the scale, but once they are, even a thousand catties cannot hold them down.

Who Huang has been involved with and how many people he has been involved with no longer matters at this point. Not having Jensen Huang matters to them.

/ 02 / Musk becomes the next "target"

Don't think that only yellow-skinned Jensen Huang is unlucky. Musk's days are not much better either.

Voting ended on November 3, 2026. The Democratic Party retook both the Senate and the House, and a bunch of deeply red states turned blue, including Texas.

For the first time since 1988, Texas elected a Democratic federal senator, and for the first time since 1994, it elected a Democratic governor.

Texas had always been the "red fortress" Musk painstakingly built. He moved the company's registration from Delaware, relocated its headquarters from California to Texas, and even built his own city, Starbase.

His calculation was simple: whatever wind blows in Washington, I don't care. Texas will always be red.

For this election, Musk spent about $474 million, setting a historical record for a non-candidate donor. Of the 37 Republican House candidates he supported, 3 won; of the 7 Senate candidates, 6 lost.

After November 3, everything changed.

The newly elected Texas governor Hinojosa had previously filed a lawsuit demanding the disclosure of more than a thousand pages of communications between Musk and former governor Abbott.

The Democrats also made their position clear: the spending cuts during Musk's leadership of the Department of Government Efficiency (DOGE)—the temporary agency he led to slash government spending—would be investigated, the use of SpaceX satellites in the Russia-Ukraine war would be investigated, and approvals for AI data centers would also be investigated.

Not just Musk. The entire AI industry saw the wind change overnight.

Reuters reported that Democratic lawmakers are already assembling a batch of regulatory bills targeting AI and data centers, preparing to push them after taking control of Congress in January 2027. Data center construction, electricity costs, model safety, legal liability, and investigations and oversight of large tech companies—not one would be left out.

Right after the election ended, Representative Yassamin Ansari posted: On the first day of the new Congress, pause advanced AI first, ban superintelligence first.

Supporting this argument is the Ban Artificial Superintelligence Act of 2026, introduced in the Senate by Sanders on September 23 and in the House the next day by Casar and 10 others. Ansari is a co-sponsor. The bill requires banning superintelligence and pausing the training of advanced models with computing power reaching 10 operations, with reckless violations by decision-makers punishable by up to 20 years in prison.

Of course, this does not mean the relevant bills will necessarily pass. Trump still holds the presidential veto power and has consistently opposed overly strict regulation of AI.

But one thing has become inevitable: "restricting AI giants" is permanently becoming a central issue in American political life.

/ 03 / A "reckoning" over wealth redistribution

Political cycles and capital cycles have always been interlocking gears. The dollar tide harvests externally, the ballot tide settles internally, and the two wheels turn alternately, round after round.

Only this time, what has been put on the table is a group of people and their companies that have seen the most concentrated wealth growth globally over the past decade.

In November 2022, on the eve of ChatGPT's release, the total market value of seven tech giants—Apple, Microsoft, Google, Amazon, Meta, Nvidia, and Broadcom—was about $7 trillion, equivalent to 27% of U.S. GDP that year.

By November 2026, the total market value of these seven companies had approached $23 trillion, equivalent to about 70% of the United States' projected full-year GDP.

In just four years, the market value of the seven tech giants increased by about $16 trillion, while U.S. nominal GDP growth over the same period was only about $6.58 trillion.

Where did the money go? The extra $16 trillion went into the pockets of a very small number of people.

According to Federal Reserve data for the second quarter of 2026: the top 10% of Americans by wealth hold about 88% of corporate stocks and mutual fund shares; the bottom 50% hold about 0.6% combined.

And who pays the price? It is left to society as a whole.

Compared with the convenience brought by AI, what ordinary people feel more directly are two things: their jobs, and the bills mailed home every month.

In an August 2026 Pew Research Center survey, 73% of Americans aged 18 to 29 believed AI would reduce U.S. jobs over the next 20 years. In 2024, that figure was 61%, a jump of 12 percentage points in two years.

A study updated by Stanford University in August this year revealed a worrying trend: AI has not yet triggered mass unemployment, but it has already begun affecting young people's job opportunities.

The study examined wage data for millions of U.S. employees. From November 2022 to June 2026, in occupations more exposed to AI, employment among those aged 22 to 25 fell by about 11%.

As if having jobs replaced by AI were not enough, the cost of living has also been rising.

PJM, one of the largest regional grid operators in the United States, covers the Mid-Atlantic and parts of the Midwest and serves more than 65 million people.

PJM has a mechanism called the "capacity market." Simply put, to ensure there is no power shortage during future peak demand periods, the grid buys future power supply capacity from generators in advance.

Now, as electricity demand from data centers surges, the grid needs to reserve more supply capacity in advance, and capacity market prices have risen accordingly. These extra costs are ultimately passed on to ordinary businesses and households through electricity bills.

PJM's independent market monitor estimated that for the three delivery years from 2025 to 2028 alone, the additional capacity market costs brought by data center demand would exceed $23 billion.

In Maryland and parts of Ohio, data center demand has even forced ordinary households to bear about $10 more in monthly electricity bills.

These grievances are being transformed into political pressure, and AI giants are also beginning to become important targets in the two-party contest. Compared with the economic accounts on paper, these political battles carrying surging public opinion are becoming the most unpredictable variable shaping the direction of the AI industry.

/ 04 / Conclusion

There is no doubt that Huang's downfall will have disastrous consequences for Nvidia and even the entire AI industry.

But clearly, the worst-case scenario is still far from arriving.

The reason is simple: compared with external threats, invisible internal reckonings are often more lethal.

As the wealth created by AI becomes more concentrated, social costs such as employment and energy may be borne by more ordinary people. When this imbalance of interests is converted into votes, tech giants will inevitably be drawn into political maneuvering, and long-tail risks will truly become greater.

This means that the odds and payoffs of Nvidia and even the entire AI industry need to be reassessed. The long-tail risks brought by political uncertainty are amplifying. Even if EPS continues to grow and the technological moat remains solid, that may not prevent further compression of valuation multiples.

After all, commercial competition has winners and losers, but political maneuvering can change the rules of the game.

Text / Aqi