Bernie Sanders Proposes $7 Trillion Public Takeover of AI Industry Through 50% Equity Tax

The legislation would create a sovereign wealth fund overseen by an independent commission, financed through a one-time 50% tax on the stock of the largest AI companies. Sanders estimates this would generate a nearly $7 trillion fund producing hundreds of billions in annual returns. Any AI firm doing $200 million in annual AI sales would fall under the scheme. The fund would then pay direct dividends to American citizens, with Sanders projecting approximately $1,000 per person annually. According to Fortune, Sanders framed the proposal as ensuring AI's economic benefits reach ordinary people rather than concentrating among billionaire founders and venture capitalists.

The mechanism is deliberately radical. Rather than regulating AI through traditional means, Sanders's approach strips private shareholders of half their equity and redirects future profits to the public. Fox Business reports the targets explicitly include OpenAI, Anthropic, and xAI, companies currently valued in the hundreds of billions with public offerings anticipated. The senator's office argues these firms built their models using creative work from millions of people without compensation, making public ownership a form of restitution.

Sanders has positioned the plan as a response to widespread public anxiety about AI's societal impact. In a Quinnipiac poll cited by his Senate office, 55% of Americans believe AI will do more harm than good, 70% expect job losses, and only 5% think AI development serves their interests. Sanders told reporters that AI and robotics will impact every American, pointing to predictions of massive job displacement, privacy erosion, and harm to children's mental health. His New York Times op-ed argued AI companies trained their models on collective creative output without paying the sources.

The political framing extends beyond economics into existential territory. Sanders's Senate website published an essay headlined AI Is a Threat to Everything the American People Hold Dear, linking unregulated AI development to risks against jobs, equality, social connection, democracy, and potentially humanity itself. This rhetoric positions public ownership not merely as redistribution but as democratic control over technology that threatens fundamental values. The proposal arrives as 74% of Americans tell pollsters the government isn't doing enough to regulate AI, suggesting Sanders is tapping genuine public demand for intervention.

AI companies have not publicly responded, though Ars Technica notes they will likely recoil at the proposal. The one-time 50% stock levy would instantly halve existing shareholders' stakes in affected companies, including employees, founders, and venture capital firms that have poured tens of billions into AI development. Companies preparing for public offerings, including OpenAI and Anthropic according to Futurism, would see their valuation narratives complicated by the prospect of government seizure.

The proposal faces virtually no chance of passage in the current Congress, where Republicans control both chambers and many Democrats maintain close ties to Silicon Valley donors. Yet its introduction matters for setting political boundaries. By anchoring the debate at public ownership, Sanders may expand what counts as moderate AI policy, potentially making less radical measures, heavier regulation or windfall taxes, appear more palatable by comparison. The bill also forces AI companies to defend their private structure publicly at a moment when antitrust sentiment runs high across both parties.

Sanders's plan assumes American citizens should exclusively benefit from AI value extraction, a framing that Action Model challenges. If AI systems are trained on global knowledge, creative works, and data from people worldwide, restricting ownership dividends to Americans raises fundamental questions about fairness and reciprocity. The proposal does not address whether foreign citizens whose data and labor contributed to AI training receive any compensation, nor whether other nations might impose mirror requirements on American AI companies operating abroad.

This tension reflects broader unresolved debates about AI's international governance. Current AI development depends on globally sourced training data, international talent, and cross multinational supply chains for chips and computing infrastructure. A unilateral American seizure of AI equity could trigger retaliatory measures, fragment AI governance along national lines, or accelerate corporate relocation to jurisdictions without such levies. The plan's nationalism sits uneasily with AI's fundamentally transnational character, suggesting any serious implementation would require international coordination Sanders has not proposed.

The bill's immediate fate is sealed by congressional arithmetic, but its longer-term influence depends on whether other politicians adopt its framework. Sanders has historically used seemingly impossible proposals, Medicare for All, free college tuition, to shift Democratic Party priorities over decades. AI policy may follow this pattern if public anxiety about the technology persists and economic inequality from AI gains becomes a salient political issue.

Democratic presidential contenders in 2028 may face pressure to address AI wealth concentration, potentially borrowing elements from Sanders's approach without embracing full public ownership. Republican counterproposals will likely emphasize American AI dominance through deregulation rather than redistribution. Meanwhile, AI companies' lobbying budgets will expand to defend against this and similar threats. The proposal's $7 trillion figure, whether accurate or not, establishes a benchmark for public claims on AI wealth that will resurface in future debates regardless of this bill's fate.