Entrepreneur and former presidential candidate Andrew Yang is pressing Congress and the public to accept a simple trade: tax artificial intelligence, and use the revenue to cut taxes on human labor. The argument, which Yang has repeated across television appearances and his own newsletter in recent months, gained fresh attention during a recent CNN segment. His pitch arrives as economists warn that payroll taxes and other labor-based levies increasingly penalize the very workforce that automation threatens to replace.
Yang’s Case for an AI Tax
During a CNN News Central appearance, Yang argued that an AI tax would broadly distribute the gains generated by artificial intelligence. He cited Anthropic CEO Dario Amodei’s public remark that AI will automate away a large share of entry-level white-collar jobs, and noted that Amodei’s proposed remedy was essentially “Tax us.” Yang contended that AI firms are already generating enormous sums of revenue, and that a portion of that wealth should flow back to the public, which Yang argues has effectively helped train the systems through the data they generate.
That framing is not new for Yang. In a late November 2025 segment on CNN’s Smerconish, he explicitly endorsed an “AI tax” or “compute tax” and repeated his attribution of the “you should tax us, the AI firms” line to Amodei. Yang argued there that policymakers need to capture “very big numbers” quickly to fund broad-based support programs, including a version of universal basic income. The consistency of his language across months of appearances signals a deliberate campaign rather than an offhand policy suggestion.
A Proposal Six Years in the Making
Yang’s AI-tax logic predates the current wave of generative AI tools. During his 2020 presidential run, he sat for an interview with The Washington Post Editorial Board and argued for shifting tax collection toward a value-added tax that could be “scaled up” on items including AI. At the time, few politicians treated automation-driven job loss as a near-term policy emergency. Yang’s VAT-plus-UBI platform was often treated as a curiosity. Six years later, with AI tools increasingly taking on tasks once done by human workers, the policy conversation has caught up to his original framing.
In his newsletter post titled “The Future Economy,” Yang continued to build the case, referencing Amodei’s call for an AI tax and arguing that the economy must adapt as technology displaces human labor. The post shows Yang promoting the AI-tax concept beyond television hits, reaching a subscriber base that skews toward tech-aware voters and policy enthusiasts. His written arguments tie the tax proposal to broader anxieties about white-collar unemployment, a concern amplified by reporting on AI-driven job displacement at companies like Anthropic.
Why the Current Tax Code Favors Machines
Yang’s proposal gains analytical weight from institutional research showing that the U.S. tax system already tilts the playing field toward capital investment and against hiring. Economists at Boston University have explained how the tax code can incentivize automation by imposing payroll taxes on labor while offering more favorable treatment to capital, software, and equipment purchases. Their analysis highlights the gap between what employers pay in taxes on a human worker versus what they effectively pay on the machines and software that replace that worker. The implication is direct: every dollar spent on payroll taxes makes a human employee relatively more expensive compared to an AI tool that faces no equivalent levy.
This structural bias means that even without deliberate policy choices favoring automation, the existing system nudges firms toward replacing workers with technology. Cutting payroll taxes or other labor-specific levies, as Yang proposes, would narrow that gap. But it would also blow a hole in federal revenue unless paired with a new source of funds, which is precisely where the AI or compute tax enters the equation.
Compute Taxes and Their Tradeoffs
The Brookings Institution has published a public finance framework examining how tax policy should adapt to an AI-driven economy. That analysis discusses compute taxes as one mechanism for countering tax-base erosion, the gradual shrinking of taxable labor income as machines take over tasks previously done by salaried workers. The Brookings framework also weighs potential downsides, including the risk that taxing computational power could slow innovation or push AI development to jurisdictions with lighter regulatory burdens.
That tension sits at the center of the debate Yang is trying to shape. A compute tax that is too aggressive could discourage the very investment that makes AI productive. A tax that is too modest would fail to generate the revenue needed to offset lost payroll tax income or fund new safety-net programs. Neither Yang nor the institutional analyses published so far have offered a specific rate or implementation timeline, a gap that leaves the proposal vulnerable to criticism from both fiscal hawks and tech advocates who want more detail before signing on.
The Regional Inequality Blind Spot
Most coverage of AI taxation focuses on the national balance sheet: how much revenue a compute tax could raise, and how that money might fund UBI or retraining programs. But there is a distributional question that gets less attention. AI development is concentrated in a handful of metro areas, primarily in the San Francisco Bay Area, Seattle, and parts of the Northeast. A tax on compute could, in theory, generate revenue from those hubs while the benefits of reduced labor taxes would flow to workers everywhere. That sounds progressive on paper.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
