As Blackstone, the world’s largest private equity firm, teams up with AI company Anthropic to automate operations across its portfolio of over 250 companies—including healthcare, financial services, and hospitality—the specter of mass layoffs is growing. The partnership, first reported by The Information, aims to slash jobs through a top-down AI overhaul, raising alarm about the societal costs of rapid automation.
But job losses are just one concern. Experts warn that AI poses a growing array of public harms, from systematic data theft and algorithmic discrimination to the concentration of economic and political power in a handful of corporations. Private shareholders, however, have little incentive to address these externalities.
A Novel Tax Proposal
In a paper published in the Columbia Journal of Tax Law, tax professor Jeremy Bearer-Friend and intellectual property professor Sarah Polcz propose a solution: require systemically important AI firms to pay a new tax with stock, not cash. The stock would be placed in a publicly managed trust, giving the public fractional ownership and limited governance powers in these companies.
“At its core, our AI tax would give the public fractional ownership, alongside private investors, in large AI firms,” the authors write. This approach, they argue, would compensate the public for the data scraped from the internet to train AI models and give citizens a voice in corporate decisions through governance tools.
Revenue and Accountability
The proposal also addresses a looming fiscal gap: as AI displaces workers, income tax revenue could plummet while demand for public services rises. The AI tax would provide a new funding source without requiring government payouts to acquire shares, unlike President Trump’s recent $9 billion gambit with Intel.
Unlike Trump’s efforts to own stakes in private enterprise, this tax would be democratically accountable. Congress would enact it, with enforceable conflict-of-interest rules and ethics requirements for managing the public trust. The fund would follow best practices from existing public investment funds, such as Calpers, which holds hundreds of billions in private company shares, and sovereign wealth funds in Norway, Abu Dhabi, and Alaska.
“There are clear precedents for publicly managed funds that own shares in private enterprise within a capitalist economy,” the authors note.
Mechanics and Timing
The equity paid would match existing share classes, avoiding the need for new securities. This aligns the Treasury with other investors who want to preserve share value. “No longer could a company turn to shareholders to say how much money it made, and then turn to the IRS and say it made nothing,” they write.
For AI companies, issuing stock is routine—tech workers often receive stock as compensation. But the authors argue that waiting to act is dangerous. Tech firms often avoid corporate income tax liability until they achieve monopoly power, as seen with Uber, Meta, and Amazon. By then, legislatures find it much harder to intervene.
“The tax system must move at the speed of the AI rollout,” they warn. “We are quite possibly headed towards a winner-take-all future where a small number of companies control ever larger swaths of our economy.”
A Fork in the Road
In one scenario, all AI gains flow to private equity and a handful of wealthy shareholders. In another, those returns are shared with workers whose jobs are being replaced. “It’s time to tax AI now before it’s too late,” the authors conclude.
The proposal comes amid broader concerns about public trust in institutions. A recent Gallup poll shows public trust in government at historic lows, and debates over AI governance are increasingly politicized. The White House has reconsidered a ban on Anthropic after its powerful new AI model drew federal interest, highlighting the stakes of the debate.
