October 28, 2026
Authors

AI systems depend on large-scale human labor throughout the AI lifecycle. Behind model development, alignment, evaluation, and deployment is a global workforce performing data annotation, reinforcement learning from human feedback, red-teaming, content review, model monitoring, and more. While AI is often framed as autonomous, these systems remain fundamentally reliant on human workers to ensure quality, reliability, and safety.

Awareness of labor rights issues in the AI data supply chain is growing, along with emerging good practices, resources, and examples for addressing them. However, companies have not sufficiently integrated these issues into their broader human rights programs, responsible sourcing policies, or AI governance efforts—and the field lacks a coherent, industry-wide approach to addressing them.

This Human Rights Assessment, conducted by BSR with a group of technology companies, examines labor and human rights impacts across the AI data supply chain and identifies actions companies can take individually and collectively to strengthen governance and responsible sourcing practices. The assessment was informed by stakeholder engagement with worker representatives, unions, civil society organizations, academics, buyers, and vendors, and is grounded in the UN Guiding Principles on Business and Human Rights.


Executive Summary Report
Human Rights Assessment of the AI Data Supply Chain cover image Executive Summary Report

Subscribe

BSR’s latest sustainability insights and events straight to your inbox.

Topics

Resilient business strategies for a complex world.

We Can Help