The training of AI systems and humanoid robots increasingly relies on a distributed workforce of gig workers operating from home in countries like Nigeria. These workers, many holding other jobs, use smartphones and ring lights to record themselves performing physical movements and actions that feed into machine learning models. Companies developing humanoids need vast amounts of real-world human motion data, and outsourcing this task to remote workers in lower-cost countries has become standard practice. However, this arrangement raises critical questions about labor standards, fair compensation, and the transparency of how AI development is actually being conducted globally.