Problem 1: Exploitation of Human Data

Foundation models were trained on web-scale text, images, voices, videos, code, and social data. Billions of people's contributions were absorbed with weak attribution, unclear consent, and little direct compensation, creating an unsustainable imbalance where creators lose incentives, platforms capture value, and legal pressure grows.