The $7 Trillion Heist
Bernie Sanders didn't come to ask. He came to take.
Last week, he dropped a legislative bomb: a one-time 50% tax on the stock of every AI company making over $200 million a year in AI revenue. Not dividends. Not philanthropy. Not a voluntary donation. A forced transfer—$7 trillion ripped from the balance sheets of OpenAI, Anthropic, xAI, and the rest, and poured into a sovereign wealth fund owned by you and me.
This isn't a policy tweak. It's a constitutional-level reordering of power. The same companies that now decide whether your job gets automated, your news feed gets manipulated, or your kid's tutor gets replaced by an algorithm are suddenly going to have half their equity stolen—and handed over to a commission that can veto their decisions.
Sanders calls it "a major, major, major step forward." I call it the first real fight worth having.
The tech elite won't like it. But neither should you.
How the Tax Actually Works
Let's cut through the jargon.
The plan isn't a corporate income tax. It's not a windfall profit tax. It's a stock seizure.
Any AI firm with $200 million or more in annual AI revenue gets hit with a one-time 50% tax on its equity. That means if OpenAI is valued at $80 billion, $40 billion vanishes from its balance sheet and lands in the new American AI Sovereign Wealth Fund. Same for Anthropic. Same for xAI, which merged with X and then SpaceX. The law requires them to split their AI business from everything else. So Musk's empire? It's now two companies: one that flies rockets, one that trains models. And only the AI half gets taxed.
The fund isn't sitting in a vault. It's invested in the same companies it now owns half of. So the government becomes a silent, voting shareholder. And here's the kicker: it doesn't just collect dividends. It can block decisions.
Your $1,000 a Year (And More)
Every American gets $1,000 a year—minimum—from the fund's 5% annual dividends. That's $4,000 for a family of four. That's a month's rent in Portland. A year's worth of groceries in rural Ohio. A down payment on a used car in Georgia.
But that's just the tip.
The rest? It's earmarked for healthcare, education, housing. Not vague promises. Not "we'll study it." Actual funding. Sanders says the fund will generate hundreds of billions annually. That's enough to cover 10 million low-income seniors' insulin. To build 200,000 affordable housing units. To eliminate community college tuition nationwide.
This isn't welfare. It's restitution.
The public didn't ask for AI to be built on their data, their labor, their attention. But we paid for it. We trained it. We clicked. We scrolled. We gave our faces to the datasets. Now we're getting paid—finally.
The Commission That Can Say No
Here's what no one else is talking about: the Independent Commission for Democratic AI.
Seven members. Nominated by the president. Confirmed by the Senate. Bipartisan by design.
They don't run the companies. They don't write code. But they hold voting shares. And they can veto any corporate decision that harms the public.
Think about that.
If OpenAI wants to launch a chatbot that can impersonate your dead grandfather and convince your grandmother to wire him $50,000? The Commission says no.
If Anthropic wants to sell its model to a foreign government for mass surveillance? The Commission says no.
If xAI merges with Tesla to create a hyper-automated, AI-driven surveillance state under Elon Musk's personal control? The Commission says no.
This isn't regulation. It's direct democracy over capital.
The CEOs Are Freaking Out
Sam Altman met with Sanders. He left "far apart." Not because he disagrees with the goal. He doesn't. He just doesn't want to give up half his company.
Sanders called out Altman and Trump-style "philanthropy" as a con. "We will give 5 percent of our profits back into the government... That's not what we're talking about," Sanders said. "That's charity. This is justice."
David Sacks, Trump's former AI czar, called it "straight up confiscation of property." He's right. It is. And that's the point.
The tech industry has spent a decade telling us that innovation requires freedom. That regulation kills progress. That billionaires are the only ones smart enough to run AI.
But what if the freedom to build AI that displaces millions of workers, manipulates elections, and erodes democracy isn't a right? What if it's a privilege—and one that's been abused?
Sacks wants "voluntary" public ownership. That's like asking a thief to return half his loot because he feels bad.
For context on how regulators are already challenging Big Tech's AI dominance, see EU Orders Meta to Open WhatsApp to Rival AI Chatbots.
Why This Will Never Pass (And Why That Doesn't Matter)
Let's be honest. This bill dies in the Republican-controlled House. Trump won't touch it. Sacks is his mouthpiece. The AI lobby has more cash than the entire Democratic caucus.
So why bother?
Because Sanders isn't trying to pass a bill.
He's trying to change the conversation.
For years, the public has been told to fear AI—to worry about robots taking jobs, about deepfakes, about bias. But we've been told nothing about who owns it. Who profits. Who decides.
Sanders is forcing us to ask: Should the people who built this technology have total control over its future? Or should the people who made it possible—through our data, our labor, our attention—have a say?
This isn't about legislation. It's about power.
And for the first time, a major political figure is saying: The power belongs to us.
The Real Enemy Isn't the Algorithm
The real enemy isn't the algorithm.
The real enemy is the idea that wealth should be concentrated in the hands of seven men in Silicon Valley.
That's the system Sanders is attacking. Not just AI. The entire logic of late-stage capitalism: that innovation is a private good, and its benefits are a gift from the elite.
This plan says: No.
AI wasn't built in a vacuum. It was built on the backs of millions of gig workers labeling images. On the data of teenagers posting selfies. On the public's trust in institutions that now sell it to the highest bidder.
We didn't just pay for AI.
We built it.
Now it's time to own it.
This isn't radical.
It's overdue.
The Real Source of AI's Power Isn't Code—It's Us
Let's be clear: no AI model woke up one day and decided to become a trillion-dollar industry.
It grew because we gave it everything.
Our search histories. Our photos. Our voice recordings. Our emotions, distilled into likes, shares, and rage-clicks. We didn't just use these platforms—we trained them. Every time you tagged a photo, every time you clicked "I'm not a robot," every time you scrolled past a lie because it made you angry—you were feeding the beast.
And now the CEOs want to act like they built this alone.
They didn't.
They just got lucky enough to be in the room when the money started flowing. The real labor? That was done by low-paid contractors in Nairobi and Manila labeling millions of images. The real data? That came from your phone, your laptop, your smart fridge.
We didn't sign up for this.
We signed up for free apps. We didn't sign up to be the unpaid workforce of a new industrial revolution.
And now we're supposed to be grateful when they offer us a free cup of coffee with our $1,000 dividend?
No.
This isn't charity. It's debt repayment.
The Split Is the Real Shock
Here's what the press isn't covering: the law doesn't just tax AI companies. It forces them to split.
Elon Musk's xAI isn't just a startup. It's a hybrid. It's part of X, part of SpaceX, part of Tesla's AI-driven surveillance network. Under Sanders' bill, that's no longer allowed. Musk has to choose: either he owns AI, or he owns rockets and cars. Not both.
Same for Google. Same for Microsoft. Same for Meta.
That's the real threat—not the $7 trillion. It's the severing of the empire.
Because these companies aren't just tech firms. They're conglomerates. And they've spent the last decade merging everything—social media, search, cloud, hardware, logistics—into one giant, unregulated machine.
Sanders isn't just taxing them.
He's breaking them up.
And that's why Sacks is screaming about "confiscation." He's not mad about the money.
He's mad about the precedent.
If you can take half of an AI company's equity, what's stopping you from taking half of its data? Its algorithms? Its patents?
This isn't about money.
It's about control.
The Commission Isn't a Bureaucracy—It's a Firewall
Let's stop calling it a "commission." That sounds like a panel of suits in D.C.
It's not.
It's a firewall.
A legal shield between the public and the most dangerous technology ever created.
Think of it like the FDA, but for AI.
You don't let a drug company sell a pill that kills people.
You don't let a car company sell a vehicle that randomly accelerates.
So why do we let a company sell an AI that manipulates elections, deepfakes politicians, or erodes mental health in teenagers?
The Commission doesn't need to understand neural nets.
It just needs to say no.
And when it says no, the company can't do it.
No exceptions.
No lobbying.
No "we'll fix it later."
That's the power we've been waiting for.
The Myth of the Benevolent Billionaire
Altman says he wants to "do good." Sacks says he supports "voluntary" ownership.
But voluntary means nothing.
It means the same thing as when a CEO says, "We're going to be ethical." Or when a bank says, "We're going to stop predatory lending."
They say it because they have to.
They don't mean it.
Because if they meant it, they'd have done it already.
They've had ten years.
Ten years to build ethical AI.
Ten years to limit surveillance.
Ten years to stop using children's data to train models.
They didn't.
Because profit is the only language they speak.
Sanders knows that.
He's not asking them to be better.
He's forcing them to be accountable.
And that's why this matters.
Not because it'll pass.
But because it's the first time a major politician has said: You don't get to own the future.
We do.