More than 1,000 economists, including 17 Nobel laureates, have signed a petition warning of large-scale job displacement from AI — and a growing number are pointing to one policy fix: tax capital, not labor.
More than 1,000 economists, including 17 Nobel laureates, have signed a petition warning of large-scale job displacement from AI — and a growing number are pointing to one policy fix: tax capital, not labor.

More than 1,000 economists, including 17 Nobel laureates, have signed a petition warning of large-scale job displacement from AI — and a growing number are pointing to one policy fix: tax capital, not labor.
The U.S. tax code taxes workers at 27 percent versus just 4 percent on new equipment, a gap economists say is increasingly indefensible as artificial intelligence threatens to permanently displace 10 million jobs.
"There's a reasonable case that with the rise in AI, it's more important to protect labor demand than to encourage more capital investment, and therefore capital taxes could be higher," said Doug Elmendorf, a Harvard University economist and former CBO director.
The effective marginal tax rate on labor stands at 27 percent, compared with 14 percent for new business assets and 4 percent for equipment, according to the Congressional Budget Office. The 2017 tax bill cut the corporate rate from 35 percent to 21 percent, and a 100 percent expensing provision for capital expenditures was made permanent last year — worth $31.8 billion to Microsoft, Amazon, Meta, Alphabet and Oracle combined in 2025, per Zion Research Group.
The capital share of national output is already at an all-time high, and AI promises to push it higher. Rebalancing taxes on capital and labor would not only slow AI-driven job replacement but also help close the largest peacetime deficits on record.
The petition, circulated this month by Stanford University's Erik Brynjolfsson, marks a notable shift among economists who have long viewed labor-saving technology as a net positive. A major survey published in March found that in a scenario where AI systems surpass humans in most cognitive and physical tasks, roughly 3 percent of the working-age population would exit the labor force by 2050, with 10 million jobs permanently lost.
Brynjolfsson argues it is in business leaders' "enlightened self interest" to complement rather than replace workers with AI. "We are right now steering them in the wrong direction," he said.
Retraining has limits
Retraining displaced workers is the most popular policy response, but it faces a fundamental problem: nobody knows what to train them for. "Five years ago, everyone was told, 'learn to code,'" said Pascual Restrepo, an economist at Yale University and a member of Anthropic's economic-advisory council. "That career advice has gone out the window. AI happens to be super good at computing. Maybe the next thing is economics. Then law. Then journalism."
Other proposals, such as wage insurance or universal basic income, compensate job losers but do not preserve employment. A "robot tax" or compute levy on AI tokens could be a nightmare to design and administer, and would discourage beneficial uses of AI such as drug discovery.
The tax system's role
The case for low capital taxes has driven U.S. policy for decades, from accelerated depreciation in 1954 to the investment tax credit in 1962. The most sweeping reductions came in the 2017 Tax Cuts and Jobs Act, which slashed the corporate rate from 35 percent to 21 percent and temporarily allowed immediate 100 percent expensing of many capital expenditures. That provision was restored and made permanent last year.
Since 2017, the marginal tax rate on new equipment has dropped from 14 percent to 4 percent, while the marginal rate on labor has edged down from 29 percent to 27 percent, per the CBO. The hyperscalers spending billions on data centers have been the biggest beneficiaries — Zion Research Group estimates the 100 percent expensing provision was worth $31.8 billion to Microsoft, Amazon, Meta, Alphabet and Oracle last year, and $50.2 billion compared with pre-2017 law.
In theory, taxing capital would slow AI deployment, though it would also deter non-AI investment. One study found the 2017 tax cuts did boost investment while raising long-run economic output by less than 1 percent.
The stakes extend beyond job preservation. The capital share of national output is at an all-time high, and AI promises to push it higher. As long as capital tax rates remain low, that limits how much new revenue flows into the Treasury — a consideration as the U.S. runs its largest peacetime deficits on record.
This article is for informational purposes only and does not constitute investment advice.