Bernie Sanders Wants Americans to Own 50% of Big AI: Socialist Shock or the First Real AI Wealth Debate?
Sanders’ proposed AI sovereign wealth fund would take a public equity stake in major AI firms. Critics call it confiscation; supporters call it payback for public data.
Bernie Sanders has placed one of the most radical AI policy ideas in mainstream politics: the American public should own a 50% stake in the largest AI companies. His proposed American AI Sovereign Wealth Fund Act would not simply tax profits. It would require major AI companies to transfer equity into a public fund, arguing that artificial intelligence is built on the collective work, data, language, images, labor and culture of ordinary people.
The proposal is designed to provoke. It does.
Supporters say Sanders is asking the question Silicon Valley wants to avoid: if AI systems were trained on humanity’s collective output, why should the profits belong only to a small group of founders, investors and tech giants? People wrote the books, created the art, posted the code, generated the conversations and built the internet. AI companies scraped, licensed or absorbed that material and converted it into trillion-dollar infrastructure. Sanders’ argument is that society deserves equity, not just promises of productivity.
Critics call it confiscation. They argue that forcing companies to hand over 50% of their stock would destroy incentives, crush innovation, trigger capital flight and turn the U.S. into a hostile environment for frontier technology. If OpenAI, Anthropic, xAI, Google or Meta believe the government can take half their future upside, investors may simply fund companies elsewhere. China would not need to beat American AI if America kneecaps its own firms.
The proposal also raises constitutional, legal and practical questions. Which companies qualify? How is valuation determined? What happens to private firms? Do foreign investors get diluted? Does the government gain voting rights? Who manages the fund? Could politicians pressure AI companies through ownership? Would the public receive dividends, services, tax reductions or social programs?
But the fact that the idea is extreme does not mean the underlying issue is imaginary. AI may create enormous productivity gains while eliminating jobs, lowering wages or concentrating power. If a handful of companies own the models, cloud infrastructure, chips and data pipelines, society may face a new form of technological feudalism. The value created by automation could flow upward while workers are told to retrain.
This is why the sovereign wealth fund concept matters. Countries such as Norway use public ownership of natural resources to fund long-term social benefit. Sanders is essentially arguing that AI is a new natural resource — not oil under the ground, but collective knowledge in the digital commons. If society created the raw material, society should share the rent.
The counterargument is that AI is not a natural resource lying untouched. Companies invested billions in compute, talent, risk and engineering. They built products. They negotiated infrastructure. They created value from chaos. If the state takes half after the fact, is that fairness or expropriation?
The debate will likely become one of the defining political fights of the AI era. The old left-right language may not be enough. This is not only socialism versus capitalism. It is a question of who owns machine intelligence when machine intelligence is trained on everyone.
The headline says Bernie wants the public to own half of AI. The deeper question is whether democracies can survive an AI boom where almost all economic power flows to a few private platforms. Sanders’ bill may never pass. But the question it raises will not disappear.