Users deserve to know when their data trains an AI.

Models are trained on nearly everything we post, often with no clear opt-out. Is transparency a basic right, or an impossible burden on AI companies?