June 24, 2026, (Inside AI) — Washington is confronting a radical proposition: the U.S. government should own a piece of the artificial intelligence industry. The debate, once confined to fringe policy circles, has surged into the mainstream as politicians and tech leaders grapple with AI’s potential to reshape employment, national security, and wealth concentration.
Senator Bernie Sanders last week proposed legislation that would compel AI companies valued at $200 billion or more to hand a 50% equity stake to a new public wealth fund. The fund, he estimates, could reach $7 trillion and pay annual dividends of $1,000 per taxpayer. Meanwhile, President Donald Trump has called for a summit with top CEOs to discuss public ownership, and Senator Elizabeth Warren is pushing for targeted taxes on server farms and corporate profits to fund social programs.
The proposals share a common premise: AI represents an economic discontinuity that demands a new social contract. Yet they diverge sharply on mechanics, and all face profound legal, financial, and practical hurdles. Despite the uncertainties, the conversation is gathering momentum—and starting it now may be the only prudent course.
The U.S. unemployment rate sits at 4.3%, showing no clear signs of AI-driven job destruction. The ultimate winners of the AI boom remain unknown. But waiting for a crisis before laying legal groundwork would repeat historical mistakes. Any attempt to shift federal revenue from workers to AI ventures would be an epochal shift, comparable to the introduction of the federal income tax in 1913.
Sanders’ plan, outlined in legislative text released last week, would create a sovereign wealth fund holding equity in the largest AI firms. A 5% dividend yield would fund direct payments to citizens. But the legislation is thin on calculations, and thresholds like the $200 billion mark could encourage firms to stay smaller or dispute their classification as AI companies. Giants like Amazon and Alphabet occupy a gray zone.
Trump’s approach is less defined. His administration has already used the Defense Production Act, Export-Import Bank loans, and the CHIPS Act to take stakes in U.S. Steel, mining firms, and Intel. Extending that model to AI would require new legislation, as federal agencies are generally barred from owning controlling equity in private firms. Peter Harrell, a former senior economic official in the Biden administration, notes that Congress would need to define which companies and activities fall under government ownership—and where the proceeds would go.
Even if a fund were established, its design would need constant evolution. Identifying an AI company will only get trickier as the technology diffuses, much like defining an internet company after the dot-com boom. The fund’s financial profile is also murky. Firms like Anthropic and OpenAI are unlikely to generate reliable taxable profits soon, meaning any fund would hold equity with uncertain returns and potential future cash calls as AI giants spend billions on chips and infrastructure.
Alaska’s $89 billion permanent fund offers a partial model. It reserves 25% of state oil revenue for future budgets and pays annual dividends averaging $1,000–$2,000 per resident. But its origins reveal a cautionary tale: after oil leases in the late 1960s, initial revenue was quickly spent by the legislature, cutting Alaskans out of the 1970s oil boom until the fund was created in 1976.
Critics argue that government stakes would socialize risk while privatizing control. Former White House AI czar David Sacks pushed back sharply on June 5, writing:
“Signs you might be trying to get your frontier AI lab nationalized: You compare it to nukes...threaten half of white-collar jobs...warn recursive self-improvement could end humanity...then race ahead anyway.”
Broader questions loom. Would the U.S. government soften regulation to protect its investments? How would foreign governments view American AI technology if profits flowed partly to U.S. taxpayers? These dilemmas might tempt policymakers to simply rely on existing corporate and income tax systems to capture any future windfall. But a fast-evolving industry does not mean government should sit back. If intervention is coming, the time to start is now.
