Ex-presidential candidate Andrew Yang says it’s time to ‘stop taxing labor’ and make AI foot the bill instead

Individual income taxes totaled $2.6 trillion in 2025, comprising more than half of U.S. federal revenue. Ex-presidential candidate Andrew Yang urged replacing taxes on labor with taxes on AI, while entrepreneur Zak Kidd proposed a per-task tax on humanoid robots; Sen. Cory Booker and Vinod Khosla have floated eliminating income tax on incomes up to $75k–$100k, though earners ≤$100k contributed only ~15% of income-tax revenue. The labor market is showing strain—unemployment ticked to 4.4% with 91,000 job losses—and several major tech layoffs have been attributed in part to AI-driven productivity gains.

Analysis

Proposals to shift tax incidence from labor to AI or to levy per‑task fees on robots materially change the marginal economics of automation rather than its headline feasibility. A per‑task charge creates a variable cost on each incremental robot deployment, favoring large incumbents who can amortize fixed robotics and integration costs across scale while deterring small/mid‑market quick swaps; this bifurcation accelerates vendor consolidation in automation hardware and systems integration over the next 12–36 months.

If policymakers tax AI usage (cloud compute, API calls) rather than capital robotics, expect a near‑term rotation from public cloud‑native architectures to either on‑prem or edge compute to avoid recurring fiscal levies — this is positive for chipmakers, data‑center capex beneficiaries, and system integrators but compresses SaaS gross margins where AI inference is a material cost. Equally important, the political timeline matters: headline proposals will spike idiosyncratic volatility, but durable legislative change requires multi‑year negotiation, meaning real balance‑sheet impact is more likely in a 1–3 year window than in the next quarter.

The market consensus that AI taxation is binary (either happens or not) misses intermediate outcomes: state‑level “task” pilots, sector carve‑outs (healthcare, defense), and exemption thresholds (small businesses) are probable and create arbitrage for firms operating across state lines or supplying modular robotics components. Finally, firms with hybrid labor+AI models (where human supervision is necessary) will have asymmetric advantages under either tax — they retain tax‑favored human payroll while extracting automation productivity, a strategic defensibility worth paying up for within reason.

Key Decisions for Investors

• Long MSFT 6–12 month call spread (buy calls / sell higher strike calls): Rationale — captures continued enterprise AI adoption and on‑prem/cloud hybrid demand if cloud usage faces marginal taxation; hedge — keep spread width to limit premium decay. Target: 15–25% upside if AI revenue growth sustains; stop-loss at 12% premium decline.
• Long XYZ (robotics/hardware supplier) 12–24 months: Rationale — a per‑task or robot subsidy/tax paradox increases demand for lower total cost of ownership robotics and larger players that integrate at scale. Risk/reward: asymmetric — adoption lag risk (12–18 months) vs 2–3x upside if humanoid/servicing robots scale; use 6–12% trailing stop.
• Short TEAM (Atlassian) 3–9 months or buy puts: Rationale — secular risk from tighter enterprise budgets and potential dampening of SaaS pricing if AI usage is taxed and on‑prem substitution accelerates; trade as a hedge against software multiple compression. Target: 20–30% downside sensitivity to 2–3 quarter slowdown; size position to offset broader tech exposure.
• Long MAR (Marriott) equity 6–18 months with call buy: Rationale — hospitality can monetise partial automation (cleaning, check‑in) even with per‑task levies because labor savings remain material; margins should improve on incremental deployments despite modest tax per task. Risk/reward: modest upside (20%+ over 12 months) with operational execution risk and potential reputational pushback; keep position size moderate.