John Iselin (Yale Budget Lab) & Ryan Nunn (Yale Budget Lab), How Potential AI Futures Would Play Out in the Current Tax System, Yale Budget Lab (July 20, 2026):
Key Takeaways
- It is widely expected that AI-induced economic growth would make federal debt more sustainable, in part through increased tax revenue, but the way such growth accrues to capital versus labor income may mitigate the revenue gains.
- We project that by 2030, in a rapid AI growth scenario . . . , federal revenues could grow by up to $216 billion, a 3.3 percent increase on top of CBO’s 2026 baseline.
- However, higher GDP growth due to AI could skew away from labor [income and toward lower-taxed capital income]. We estimate that—in the rapid growth scenario—total [tax] revenue gains from AI would be roughly twice as large if we held the capital and labor shares fixed at their 2026 levels. . . .
From the conclusion:
Unfortunately, the additional output growth that is often expected from AI—a clear positive for the fiscal picture—is not the only potential impact that matters. Another widely expected consequence of AI adoption is a reduction in the labor share of income. This bias in AI-induced growth reduces the revenue increase one would otherwise expect, given the preferential tax treatment afforded to capital income. Further, any labor inequality effects of AI could increase revenues (in the case of increasing inequality) or reduce them (in the case of decreasing inequality).
Because these channels are of first-order importance for understanding AI effects on tax revenues, our analysis focuses on their roles. . . .
Related TaxProf Blog coverage:
- Growiec, Prettner & Szkróbka: Workers’ Incentives and the Optimal Taxation Of AI (July 18, 2026)
- NY Times: How to Tax AI (June 14, 2026)
- Thorndike: The Forgotten Ancestor of AI Taxes (July 7, 2026)
- Bearer-Friend & Polcz on Taxing AI and the American AI Sovereign Wealth Fund Act (June 9, 2026)
- Bearer-Friend & Polcz Op-Ed: Let The Public Share In The Rewards Of Artificial Intelligence (May 11, 2026)
- Podcasts: Bearer-Friend & Polcz on Taxing AI (Apr. 20, 2026)
- SSRN Review & Roundup: Eyal Reviews Harpaz’s Taxing AI (Mar. 6, 2026)
- Harpaz: Taxing AI (Mar. 3, 2026)
