Rapid adoption of artificial intelligence could grow U.S. federal tax revenues by up to $216 billion in 2030, yet that figure would be twice as large if income gains weren't skewed toward capital instead of labor, the Yale Budget Lab said.

The potential gains in U.S. tax revenue scale up as income gains accrue toward labor income, although the aforementioned high-end estimate is paired with higher labor income inequality, or more unequal distribution of income among workers, according to a report published Monday. Tax revenues would be lower if AI gains fed more into capital income because of the more generous exemptions and lower rates of capital taxation compared with labor taxation, the report said.

Under the rapid adoption scenario, the U.S. gross domestic product could rise by more than 7% above the Congressional Budget Office's baseline for 2025 through 2030, yet labor income would fall by nearly 1% and capital income would rise by more than 17%, according to the report.

If AI is adopted at a moderate pace, 2030 tax revenues could rise by as much as $127 billion, and they would barely change at a slow adoption rate, the report said. In almost every scenario, the tax gains are driven mostly by capital taxes and corporate income taxes, with taxes on wages generally falling, according to the report. The exception is under the moderate adoption scenario, where labor income tax gains would generally rise.

AI presents a double-edged sword for tax compliance efforts, potentially making tax avoidance both easier to detect and more accessible to taxpayers, the report said.

--Editing by Neil Cohen.

For a reprint of this article, please contact reprints@law360.com.