ITEP has a new brief out today looking at a question that is likely to become increasingly important as AI transforms the economy: What role should federal tax policy play in responding to AI?
The report doesn’t endorse a single “AI tax.” Instead, it lays out three broad ways Congress could use the tax code as AI creates enormous new wealth while potentially displacing workers and imposing new costs on communities:
- Make sure the economic gains from AI are shared more broadly, including by strengthening taxes on wealthy investors and large corporations.
- Tax some of the specific costs associated with AI, with potential approaches ranging from taxes tied to job displacement to taxes on AI usage or the enormous energy demands of data centers.
- Give the public a direct stake in AI’s growth, including through proposals that would require large AI companies to provide stock to a publicly owned investment fund.
One especially relevant piece of the analysis looks at how the tax code already tilts the playing field toward automation. Under current law, a company that spends $500,000 on a machine can generally deduct that investment immediately, while a company paying a worker $500,000 over 10 years deducts those wages over 10 years. As AI makes it easier for companies to replace some forms of human labor, those tax incentives could become much more consequential.
The report also digs into some of the practical challenges with newer ideas, including taxes on companies that replace workers with AI, taxes on AI “tokens,” energy taxes on data centers, and proposals for public ownership stakes in major AI companies.
The larger point: Policymakers don’t have to treat the economic consequences of AI as something that simply happens to us. The tax code is one of the major tools Congress has to influence who benefits from AI, and who bears its costs.
