A PYMNTS Company

A New Idea Is Gaining Ground: Tax AI Computing Power to Offset Job Losses 

By  |  May 7, 2026
AI, funding, investments

Artificial intelligence is expected to eliminate millions of jobs, shift enormous amounts of wealth away from workers and concentrate economic power in the hands of a small number of companies. Policymakers do not yet have a clear answer for what to do about that. But one idea is starting to move from the fringes into mainstream economic debate: tax the computing power that makes AI run.

    Get the Full Story

    Complete the form to unlock this article and enjoy unlimited free access to all PYMNTS content — no additional logins required.

    yesSubscribe to our daily newsletter, PYMNTS Today.

    By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions.

    The Wall Street Journal reports that a growing number of economists, investors and policy advocates are taking a serious look at what is being called a “compute tax,” a levy on the processing power used to train and run AI systems. The idea has picked up steam in recent months as concerns about AI-driven job displacement have intensified.

    “Half a year ago, it was something you would hear about only in very select circles,” Anton Korinek, an economics professor on leave from the University of Virginia Darden School of Business, told the Journal. “It really has become much more mainstream in the last three months.”

    The concept is not entirely new. Bill Gates floated a similar idea in 2017, proposing a tax on robots to compensate for the labor they displaced. The current version targets AI infrastructure more directly. Under one model, companies that operate data centers would pay the tax. Under another, businesses and consumers would be taxed based on their consumption of AI processing, measured in tokens, the basic unit AI systems use to generate output.

    Former presidential candidate Andrew Yang, a longtime supporter of universal basic income, argued in the Journal that the biggest AI companies are not paying taxes proportionate to the wealth they are generating and absorbing. He said any new revenue from an AI tax should flow directly to ordinary people, not disappear into general government coffers.

    Not everyone agrees. Critics say a compute tax is too blunt an instrument. Pascual Restrepo, an associate professor of economics at Yale, pointed out to the Journal that AI is already being used in drug discovery, weather forecasting and fraud detection. Raising the cost of computing power across the board would slow those applications too. He also warned that heavy taxation could push U.S. AI development overseas.

    Some economists argue there are better tools available. Erik Brynjolfsson, director of the Stanford Digital Economy Lab, told the Journal that the current U.S. tax structure already creates a problem: companies that replace workers with machines pay less in total taxes, including payroll taxes, than companies that employ the same number of people. Fixing that imbalance, he suggested, could be more effective than a new tax on computing.

    OpenAI CEO Sam Altman, who once supported universal basic income as a response to AI-driven economic disruption, recently told The Atlantic he no longer holds that view as strongly as he once did.

    The debate is expected to grow louder as the 2026 election cycle heats up. Public anxiety about AI and job security is rising, and economists told the Journal that politicians will face increasing pressure to offer concrete responses. Whether that leads to a compute tax, a restructuring of existing tax policy or something else entirely remains an open question, but the conversation is no longer confined to academic circles.