OP-ED: Bernie Sanders Wants to Nationalize Half of the Artificial Intelligence Giants

A Carefully Calibrated and Gradual Form of Public Participation

Sanders’ proposal is not nationalization in the Soviet sense of the term. The senator is careful to make this clear. He is not calling for the government to run the labs, hire the engineers, or decide on model architectures. He is calling for 50% of the shares in companies whose market capitalization exceeds a certain threshold to be transferred to a democratic sovereign wealth fund—similar to the Norwegian Government Pension Fund—whose returns would fund a partial universal basic income, a fund for worker retraining, and a public infrastructure investment program. The threshold mentioned in the op-ed is around $50 billion in market capitalization. This would directly target OpenAI, Anthropic, xAI, as well as the AI divisions of Microsoft, Google, Meta, and Amazon.

The proposed mechanism would be phased in gradually. Companies would have the choice between voluntarily transferring shares to the public fund in exchange for tax breaks and easier access to subsidized computing infrastructure, or facing heavy taxation on undistributed profits. Sanders explicitly cites the Norwegian model, which today holds approximately 1.5% of all publicly traded shares worldwide thanks to North Sea oil revenues, and which helps fund public pensions without placing a burden on assets. The senator believes that artificial intelligence is to the United States what oil was to Norway in the 1970s: a transformative resource whose collective ownership will determine social justice for future generations. Oil came out of the ground. AI comes out of our writings, our voices, our faces. The analogy is deliberately stark.

I find the comparison with Norway more powerful than it appears. A country of five million people transformed a geological windfall into a sustainable safety net. The United States holds a far greater resource in its hands, and the current trajectory resembles organized squandering.

The social context that makes the proposal resonate

Sanders’ op-ed isn’t falling on deaf ears. It comes after several months of announcements of massive layoffs in tech, publishing, translation, and customer service. IBM announced in late 2025 that it would replace 7,800 administrative positions with AI systems. Klarna had already admitted in 2024 that its chatbots were replacing the equivalent of 700 full-time jobs. The World Economic Forum’s “Future of Jobs 2025” report estimates that 92 million jobs will be displaced globally by 2030, even if some will be recreated elsewhere. The pain is no longer theoretical. It’s in the inboxes of proofreaders, in the hallways of law firms, in animation studios, and in call centers in Manila, Casablanca, and Dublin. Sanders speaks to these people. He doesn’t promise them that technology will retreat. He promises them that they won’t be alone in facing it.

The proposal also resonates with a more unexpected audience. A segment of working-class Republican voters shares this diffuse anger toward a tech elite perceived as out of touch. A 2025 Pew Research poll shows that 57% of Americans believe AI companies have too much political power, and 62% think the benefits of AI should be better redistributed. This mistrust cuts across party lines. By proposing a concrete mechanism, Sanders has caught the political establishment off guard—an establishment that had previously been content to hold congressional hearings that led nowhere. He’s forcing the conversation to come down to earth. Not abstract power. Concrete ownership.

What strikes me is the weariness behind the anger. People aren’t asking for the moon. They’re asking for an explanation of why their jobs are disappearing in an equation in which they’ve never had a stake. Sanders’ question brings that weariness to the table.

The Case Against Constitutional and Economic Unrealism

Objections were not long in coming. In the hours following the proposal’s release, several voices spoke out, calling it unconstitutional, economically suicidal, or politically unfeasible. Legal scholars at the Cato Institute pointed out that the Fifth Amendment protects private property and requires just compensation for any expropriation. Economists at the Hoover Institution warned that such a measure would trigger a flight of capital and talent to the United Arab Emirates, Singapore, or China, where this type of proposal would be unthinkable against national champions. Editorialists at The Wall Street Journal have described it as an ideology hostile to innovation. Sanders knows these criticisms by heart. He has been hearing them since 1981. He responds by pointing out that compensation is specifically built into his plan, in the form of an exchange: shares in exchange for tax breaks, access to infrastructure, and contractual guarantees.

Other objections are more serious. How can the public share of the collective contribution be assessed? How can we prevent the fund from becoming a political tool manipulated by the administration in power? How can we guarantee the scientific independence of laboratories if half of their capital is in the hands of a public entity? Sanders points to the Norwegian model, where the fund is managed by an independent commission with strict ethical rules, and where voting at general meetings is regulated to prevent any operational interference. Is this response sufficient? Probably not to convince the most skeptical. But it has the merit of existing, and of transforming moral outrage into a debatable legal framework. This is precisely what makes the op-ed dangerous for its opponents. It does not remain merely rhetorical. It lays out a concrete mechanism.

What strikes me is the asymmetry of the debate. Sanders is criticized for proposing something imperfect, while a status quo that transfers a thousand billion to a handful of investment funds is accepted without a murmur. The imperfection of a solution never negates the gravity of the problem.

From VAT to patents, the government has always been a silent shareholder

To ground his proposal in American history, Sanders cites several precedents. He points out that the federal government has long been an active investor in strategic sectors. The Apollo program funded technologies whose civilian spin-offs fueled prosperity in the decades that followed. The National Institutes of Health funds biomedical research that pharmaceutical companies leverage to develop drugs sold at exorbitant prices. The internet itself originated from ARPANET, a military project. Sanders argues that in all these cases, the government bore the initial risk without subsequently claiming its share of the profits. The current proposal aims to correct this historical anomaly and enshrine in law what should have been established long ago: whoever finances the risk must receive a share of the returns.

Economist Mariana Mazzucato, implicitly cited in the op-ed, has been advocating this thesis for years in her books The Entrepreneurial State and The Value of Everything. She demonstrates that the major innovations of contemporary capitalism—from touchscreens to recommendation algorithms—received massive public support before being monetized by the private sector. Sanders takes the argument to its political conclusion. If the state is de facto a giant venture capitalist, it must cease to be an invisible one. It must receive dividends, and those dividends must fund public services. The logic is almost banal when put this way. It is precisely this banality that is unsettling. It reveals the scale of the current anomaly. The American taxpayer is the worst-paid investor in the history of capitalism.

I often tell myself that the most powerful arguments aren’t the most original ones. They’re the ones that put a name to a reality that everyone sensed but couldn’t quite articulate. Sanders isn’t discovering anything new. He’s putting a clear label on a mechanism that people were trying to keep vague.

A Proposal That Extends Beyond the United States

Sanders’ op-ed is not limited to an internal U.S. debate. It sends ripples through European capitals, which have been grappling since 2024 with the implementation of the AI Act and struggling to produce homegrown champions to rival the American giants. In Paris, the impact is palpable. The French government has been talking for years about digital sovereignty and strategic public investments through the Public Investment Bank. Sanders’ proposal could serve as a catalyst to reignite the debate on public investment in Mistral AI or the creation of a dedicated European fund. In Berlin, where the Sondervermögen has demonstrated that political will can mobilize hundreds of billions in just a few months, the idea of a sovereign technology fund is gaining traction in certain social-democratic circles.

In Beijing, the situation is different but instructive. Through golden shares and indirect stakes, Chinese authorities already hold a significant share of the capital of national tech giants such as Alibaba, Tencent, and Baidu. The Chinese model is obviously not democratic in the sense that Sanders understands it, but it illustrates one thing: partial public ownership of tech giants is not a utopian fantasy. It exists. It is even the norm in economies that take technological sovereignty seriously. The United States has become the exception, not the rule. Sanders turns the mirror around and confronts his country with its own singularity. This singularity has long been presented as an advantage. The senator now presents it as a vulnerability, a gaping hole in the social contract, a flaw that widens with every stock market quarter.

There is a cruel irony in the situation. The United States has long mocked European and Chinese models for their interventionism. Today, it is the United States that is watching its own public infrastructure crumble while ten companies divide up the new economy among themselves. History sometimes has a cold sense of humor.

A lost cause, perhaps, but one that shifts the boundaries

Let’s be realistic. Given the current political makeup of Congress, Sanders’ proposal has no chance of passing. The Senate is dominated by a majority hostile to any form of public participation in the private sector. The House of Representatives is rife with tech interests heavily funded by Silicon Valley lobbyists. Contributions from Sam Altman, Marc Andreessen, Peter Thiel, and Elon Musk to the 2024 and 2026 election campaigns total more than $250 million. The financial clout of the opponents is overwhelming. Sanders knows this. He makes no secret of it. His op-ed is not a bill ready for a vote. It is an ideological marker, a reference point intended to reconfigure the boundaries of what can be said.

The history of political ideas shows that radical proposals often serve to make moderate proposals negotiable. Sanders’s “Medicare for All” in 2016 was never adopted, but it profoundly transformed Democratic platforms all the way up to the Inflation Reduction Act of 2022. The Green New Deal was never passed, but it shaped the entire American climate debate. The proposal on AI follows the same logic. It won’t be passed in 2026. But it opens up the possibility—in two, five, or ten years—that a more moderate version—a public fund of 10% or 15%, or a tax on AI profits earmarked for a partial universal basic income—might become politically acceptable. Sanders isn’t playing the electoral cycle. He’s playing the generational cycle.

I find a certain nobility in this approach to politics: accepting defeat now in order to win later; putting forward an idea that you know is doomed in the short term, because you know it will take root once conditions have changed. It’s slow, it’s thankless, and it’s rarely spectacular. Yet that’s how societies truly change.

A Map of Support and Awkward Silences

Several voices quickly rallied behind Sanders. Economist Joseph Stiglitz, the 2001 Nobel laureate, hailed the proposal as courageous and conceptually sound. Robert Reich, former Secretary of Labor under Clinton, published an op-ed supporting the idea of a citizen’s dividend on AI. Senator Elizabeth Warren indicated that she would carefully examine the proposed mechanisms. On the labor front, the AFL-CIO issued a statement calling for an urgent national debate on the ownership of automation technologies. Organizations representing artists, such as the Authors Guild and the Writers Guild of America, see the proposal as a logical extension of the legal battles they are waging against the unpaid use of their works in training models.

The silences speak volumes. The official Democratic Party remains cautious, aware of the tech funding that powers its own infrastructure. Mainstream editorialists have overwhelmingly adopted a “realistic” angle—a polite way of sidestepping the substantive debate. Tech executives, for their part, have retreated into strategic silence. Sam Altman has not commented publicly. Dario Amodei, founder of Anthropic, discussed the challenges of redistribution in a podcast without mentioning Sanders. Sundar Pichai and Satya Nadella have preferred to focus on their product announcements. This silence is rational. Any direct response would risk amplifying the reach of a proposal that the AI giants would prefer to see fade away amid media indifference. Silence is the weapon of choice for the powerful when the opposing argument is valid.

This silence speaks to me more than any statement. When an idea is weak, it is publicly crushed. When an idea is dangerous, it is ignored so that it withers away. The giants’ silence in the face of Sanders looks like a strategy. It could very well end up backfiring on them.

What Will Remain of This Op-Ed in Ten Years

Bernie Sanders’ op-ed will not change U.S. law in 2026. It will not lead to nationalization. It will not force OpenAI or Anthropic to back down. But it will have achieved something more subtle and enduring. It will have clearly and forcefully articulated the question that all democratic governments will have to confront in the years to come. Who owns the wealth generated by artificial intelligence? The question has been posed. It can no longer be brushed aside. It will resurface with every productivity report, every wave of layoffs, every lawsuit filed by rights holders, and every time a new data center opens, devouring the water and electricity of an entire region. It will resurface in the statements of labor unions, in election platforms, in European negotiations, and in UN forums. Sanders did not invent the problem. He gave it a form clear enough for it to become a political issue.

The proposal for 50% public ownership will likely be renegotiated, watered down, and reworked. It may end up being 20%, 15%, or 10%. It will take the form of a specific tax, a transition fund, or a citizen’s dividend. It will vary depending on the country, political culture, and balance of power. But the framework has been established. And in ten years, when employment statistics have been reshaped by autonomous agents, when inequalities have reached levels that even orthodox economists will refuse to defend, the op-ed of June 1, 2026, will be seen as a tipping point. Not a victory. Not a defeat. A milestone. The moment when someone, from Vermont, said what others dared not put into words. Artificial intelligence is too important to belong to just ten people.

I close this op-ed with an image in mind: an elderly man with white hair, sitting in a modest office, typing on a keyboard an idea that billions of dollars will try to stifle. He is 84 years old. He knows he may not live to see the outcome of his fight. Yet he writes anyway. There is something about this quiet determination that resembles a definition of political courage. Not the kind of courage that shouts. The kind that lays an idea on the table and walks away, knowing that others will pick it up.

By Jacques Pj Provost, columnist