AI's Data Labor Supply Chains Leave Workers Without Recourse

When Meta ended its Nairobi data-labeling engagement, over 1,000 workers employed by Sama, a San Francisco-based firm that supplied AI trainers for Meta's systems, were laid off. For a worker named Grace, that meant losing a job that paid 30,000 Kenyan shillings (approximately $230) a month before taxes, in a workplace where Meta's tracking system recorded bathroom breaks to daily productivity levels. The story, reported by Tech Policy Press, is a case study in how outsourced AI labor supply chains create accountability gaps that workers end up absorbing.

The architecture of the problem is structural, not incidental. According to SOMO's March report, almost 500 companies are currently active in AI data collection and labeling, and Amazon, Google, Meta, Microsoft, and Nvidia combined use at least 30 intermediary companies for data work. That many contracting layers means, as the article puts it, responsibility is so diluted that it diffuses across companies and regulators.

Joan Kinyua, president of the Data Labelers' Association, captures the accountability gap plainly: "If I was working directly for Meta, I can tell you for free, these are not the challenges I'm going to speak about." Workers doing foundational annotation work have no direct relationship with the companies whose AI systems depend on their labor.

The scale of what is at stake is growing. Global Insight Services predicts the data annotation market will expand from $1.2 billion in 2024 to $10.2 billion by 2034. Without changes to how contracting responsibility is structured, more market growth plausibly means more workers in precarious positions without direct recourse. What the reporting does not resolve is whether any of the named companies are actively piloting accountability reforms, or whether regulatory frameworks are developing quickly enough to keep pace with that growth.

Worker associations like the Data Labelers' Association represent an emerging point of leverage. The kind of reputational scrutiny that reshaped garment supply chains may ultimately matter as much as formal regulation here, and SOMO's mapping of hundreds of active companies gives researchers and policymakers a documented foundation to work from.