Everyone Wants to Tax AI. Almost No One Agrees How.
Proposals to capture some of AI's productivity windfall are multiplying in Washington and abroad, but the technical and political design questions are still wide open.
The idea of taxing artificial intelligence has moved from think-tank panels to active legislative drafting on both sides of the Atlantic. The harder question — what exactly to tax, and how — remains unresolved.
Three broad approaches dominate the debate. The first would tax the inputs: compute, training data, or imported chips, the way some jurisdictions tax energy or specific industrial activities. The second would target the outputs: a per-query or per-token levy on commercial AI services, modeled loosely on excise taxes. The third would aim at the displaced wages, taxing companies in proportion to the labor income their AI deployments replace.
Continue reading
3 more paragraphs for subscribers.
Free readers get a daily preview and the weekly long-form. Premium unlocks every brief in full, AM & PM editions, alerts, and the archive.