Summary
Joint report from Alphabet Workers Union-CWA and TechEquity documenting working conditions of U.S.-based AI data workers — challenging the assumption that labor exploitation in AI is exclusively a Global South problem. 86% of surveyed workers worry about meeting financial needs; median wage is $15/hour for 29-hour workweeks ($22,620 annual). 25% rely on public assistance. Only 23% have employer-provided health insurance.
Key Points
- 86% of U.S. AI data workers struggle to meet financial needs
- Median wage: $15/hour; median hours: 29/week; median annual income: ~$22,620
- 25% receive public assistance (SNAP, Medicaid, housing assistance)
- 23% have employer-provided health insurance
- 87% are regularly assigned tasks they haven’t been trained for
- Workers face arbitrary account suspension without explanation or appeal process
- Algorithmic management: work quantity and quality monitored by automated systems; humans managed by code
- Finding challenges conventional wisdom: AI data work exploitation isn’t limited to the Global South
Newsletter Angles
- The domestic hidden labor story: Most AI labor coverage focuses on Kenya and Venezuela. This report documents the same exploitation happening in the U.S., at $15/hour, under algorithmic management with no job security.
- Public assistance subsidizing AI: AI data workers earning $22,620/year and receiving SNAP benefits means the government is effectively subsidizing AI company labor costs. The social safety net is the margin that makes underpaid ghost work economically viable for workers.
- Alphabet Workers Union framing: The AWU-CWA producing this report matters — they’re building the case for labor organizing within the AI supply chain. The report isn’t neutral research; it’s movement infrastructure.
Entities Mentioned
- Amazon — AMT named as primary platform; broader data work ecosystem
- Mechanical Turk Pattern — U.S. domestic dimension of the concealment pattern
Concepts Mentioned
- Mechanical Turk Pattern — U.S. workers facing same structural conditions as Global South counterparts
- Leverage Erasure Through Automation — algorithmic management as a form of leverage erasure; workers managed by code
Notes
Source bias: Produced by Alphabet Workers Union-CWA — a labor organization with a clear advocacy position. Data collection and framing serve labor organizing goals. Cross-reference with independent labor research for verification. The directional findings (low wages, poor conditions, algorithmic management) align with independent research; specific percentages should be treated as union survey data.