The piece further emphasizes that the use of AI products implicates consumers in this global network of exploitation, as tech companies often hide the reality of the poorly paid and grueling human labor that trains and manages their products. The article concludes with a personal account of a data annotator in Uganda, underscoring the lack of job security and opportunities, and the constant fear of being laid off.
Key takeaways:
- Content moderators and data annotators in countries like Kenya and Uganda are subjected to intense, psychologically damaging work conditions, often having to review violent, explicit, and disturbing content for companies like Meta.
- These workers are expected to meet stringent performance targets, with little time to process the traumatic content they witness. Many report long-term psychological damage and some have even attempted suicide.
- The majority of these workers are on insecure, short-term contracts and work in fear of losing their jobs. They are also subject to oppressive surveillance and are often discouraged from forming trade unions.
- The global marketplace for data annotation is booming, expected to be worth over $17bn in 2030. However, the exploitation of workers in the global south continues, as tech companies take advantage of high unemployment rates and the lack of well-paid jobs with employment protections.