News Details

Gig workers selling personal data for AI training are sensing opportunity but in reality, are they making themselves obsolete. All your AI-powered chatbots like ChatGPT and Claude depend on data for training. However, research suggests these models will run out of content by mid-2026. This is where workers like Priyanka (23) and Laveena (19), based in Ranchi and Kanpur respectively, come in. They document their everyday lives, sharing photographs, audio clips, and video clips to apps like Kled AI or Neon Mobile, which pay contributors for uploading their data to train AI models. Similarly, Nikumbo Dondangi (27) in Kenya records audio clips and ambient sounds, selling them to Silencio for AI training. Workers earn around $10 for a video or over $100 a month. However, these platforms offer more than just monetary compensation—they provide a license for data use worldwide, exclusive, irrevocable, transferable, and royalty-free. A single 20-minute video can train a chatbot for many years, and photographs can be used in facial recognition software without consent. Videos and audio clips can be repurposed into deepfakes or advertisements globally. Workers’ consent is vague; they agree to terms but have no control over data usage once it enters AI systems. Current laws do not protect them from misuse. Gig economies, including this one, follow a pattern of low, irregular income with no protections. With global demand for data, prices have dropped, making data sales a diminishing return. The paradox is that the data workers sell will make them obsolete in the future. The question remains: Who truly owns AI built on human lives? Workers see it as opportunity, but platforms gain long-term benefits while workers receive one-time payments. The AI revolution involves transferring human identity, memory, and behavior into machine systems, leaving little room for reversal.