Bernie and Trump Agree the Public Should Own AI. The Companies Lose Billions.

For anyone told the public will finally share in the AI boom: what Bernie, Trump, and Sam Altman actually proposed, why the math looks more like a bailout than a windfall, and who really gets paid.

Image by Ahmed M. Abdelfattah

In the same week, a self-described democratic socialist and a Republican president found themselves arguing for the same thing.

Bernie Sanders proposed that the public take a 50 percent stake in the largest AI companies. Days later, aboard Air Force One, Donald Trump floated the government taking direct equity in OpenAI, Anthropic, and xAI, and called it a beautiful thing. Sam Altman, whose company they were both discussing, said he agreed with the general idea.

Everyone, suddenly, wants the public to own a piece of AI.

Almost no one mentioned the part that should change how you read all of it: these companies are losing billions of dollars a year.

That omission is the story. Because there is a world of difference between the public owning a slice of something that prints money and the public being handed a slice of something that burns it. And the words being used, “wealth fund,” “partnership,” “the people benefit,” are doing everything they can to keep you from noticing which one this is.

Bernie’s 50 Percent, Altman’s Hour, Trump’s “Beautiful Thing”

Start with the facts, because the convergence is real and it is genuinely strange.

In early June, Sanders unveiled what he called the American AI Sovereign Wealth Fund Act. As reported by Fortune, it would impose a one-time 50 percent equity tax on the stock of the largest AI firms, paid in shares rather than cash, with the proceeds funneled into a public wealth fund meant to spread the fortune those companies generate.

Soon after, Altman asked to meet Sanders, and the two spoke for nearly an hour in the senator’s office. According to people familiar with the conversation, Altman told Sanders that he, too, wants the public to hold equity in AI companies. He would not go to 50 percent, but he wanted to work together on the general idea.

Then came Trump. Speaking to reporters on Air Force One on June 5, he described an arrangement where, in his words, the American people could benefit from the success of AI. “You make them a partnership in this revolution,” he said. “It would be a beautiful thing.”

This is not idle talk from his administration. It took a 10 percent stake in Intel last year and has floated direct involvement in other troubled companies. The idea of the government on the cap table is not hypothetical for this White House. It is a habit.

Then I Looked at the Balance Sheets

I will admit my first reaction was that this might be good.

For years the criticism of the AI boom has been that a handful of companies capture all the upside while the public absorbs the costs: the energy bills, the water, the jobs, the disruption. A public stake sounds like the correction. If the public is bearing the risk, the public should share the reward.

Sanders has made exactly that argument for years, and there is principle in it.

Then I looked at the balance sheets, and the picture inverted. The thing being generously offered to the public is not a share of a fortune. It is a share of a deficit.

OpenAI Loses $14 Billion. xAI Spends $2 for Every $1.

Here is what these companies actually do with money.

According to internal projections reported by The Information, OpenAI is on track to lose around 14 billion dollars in 2026 alone. Elon Musk’s xAI lost 6.4 billion dollars from operations last year on just 3.2 billion in revenue, which means it spent two dollars for every dollar it took in, and the gap is widening.

Anthropic is the lone partial exception, on pace for a profitable quarter with annualized revenue approaching the tens of billions. But even that profitability is fragile, built atop training costs that have already consumed billions.

Image by Ahmed M. Abdelfattah

So picture the deal on the table. The public is invited to take an ownership position in a set of companies that, with one shaky exception, lose enormous sums every single year, with no clear date at which that stops.

A 50 percent stake in OpenAI today is a 50 percent claim on its losses as much as its someday-maybe profits.

You do not build a sovereign wealth fund out of that. You build something else, and it has a different name.

Norway’s Fund Owns Oil. This One Owns Losses.

The phrase Sanders chose, sovereign wealth fund, points to real models, and the comparison is exactly where the proposal falls apart.

Norway’s fund and Alaska’s permanent fund, the templates everyone reaches for, work because they own profitable things. Oil generates cash, the fund collects it, and the public gets a dividend. The asset pays you for holding it. That is what makes it a wealth fund.

A stake in a company losing 14 billion dollars a year does the opposite. It pays no dividend, because there is no profit to distribute. It carries a continuing obligation, because losses have to be funded by someone. And it asks the public to hold the downside of an unproven bet while hoping a future payoff materializes.

There is a word for putting public money into a money-losing enterprise to keep it going, and it is not investment. It is a bailout.

The test is simple enough to apply to any version of this idea you hear in the coming months. Is the asset making money or burning it? If it is burning it, you are not being offered a fortune. You are being asked to cover a fire.

Insiders Sell at the Top as the Public Is Invited In

Now add the calendar, because it answers the question of who this actually serves.

This burst of public-ownership enthusiasm arrives at the precise moment the people who already own these companies are about to convert their stakes into cash.

SpaceX, which houses xAI, is set to price what would be the largest IPO in history this week. OpenAI is finalizing its own IPO paperwork. An initial public offering is, fundamentally, the event where early insiders and investors sell.

So the timeline reads: just as the founders and venture backers prepare to cash out at trillion-dollar valuations, the political class discovers an urgent interest in the public buying in.

Image by Ahmed M. Abdelfattah

I am not claiming a conspiracy, and the dates may simply have collided. But the sequence is worth saying plainly. The insiders sell high, and the public is encouraged to buy into the losses they are leaving behind, with the whole thing wrapped in the language of fairness.

That is the hinge on which “partnership” turns into something less flattering. A partnership in the profits is one thing. A partnership that socializes the losses while the original owners take their gains and leave is the oldest move in finance, and it has been dressed in populist clothing from both ends of the spectrum at once.

If AI Pays Off, Early and Cheap Wins

The proponents have a real argument, and it deserves its best form, not a strawman.

If AI turns out to be as transformative and as profitable as its champions insist, then getting the public an ownership position now, while the companies are still unprofitable and the entry is cheap, could pay off enormously later. Exactly the kind of early, patient, contrarian bet that private investors are being richly rewarded for making today. Why should only venture capital get that upside?

There is also a governance argument: a public stake could give citizens real leverage over companies that are reshaping society, a seat at the table rather than a complaint from outside it.

And Sanders’s core instinct, that the public funded much of the foundational research and is absorbing the boom’s costs, so the public deserves a claim on its rewards, is not unreasonable.

Each of those holds, and together they still do not resolve the central problem.

The early-bet argument works for private investors because they choose the risk with their own money and can afford to lose it. Forcing that speculative risk onto the public balance sheet is a different proposition, especially when the downside is immediate and the upside is a hope.

The governance goal can be achieved through regulation and antitrust without putting taxpayers on the cap table of firms bleeding cash. And the fairness argument only delivers fairness if there is a reward to share, rather than a capital bonfire to subsidize.

The principled version of this idea is coherent. What is being floated this week, in the language of beauty and partnership, is not yet that. It is a stake in losses, offered at the moment the winners are heading for the exit.

The White House Meeting Is the Tell

The executives are due at the White House within days, and that meeting is the tell.

Image by Ahmed M. Abdelfattah

Watch whether “public ownership” hardens into actual legislation with actual terms, or stays a slogan that everyone endorses and no one defines. Vagueness is what lets a bailout pass as a windfall.

Watch the IPO calendar against the policy talk, because the closer the two run, the clearer it is who the partnership is really for.

And watch the word “fund,” because the moment anyone has to explain how a fund pays the public from companies that earn no profit, the illusion gets harder to maintain.

The remarkable thing about this week is that the argument over AI has finally reached the question of ownership, which is the right question and a sign of how seriously the technology is now taken.

The unremarkable thing, the part as old as money itself, is that the people being offered a piece may be getting a bill.

Everyone agrees the public should own AI. No one has yet answered whether the public is being handed a fortune or asked to cover a loss.

This story is published on Generative AI. Connect with us on LinkedIn and follow Zeniteq to stay in the loop with the latest AI stories.

Subscribe to our newsletter and YouTube channel to stay updated with the latest news and updates on generative AI. Let’s shape the future of AI together!