Model Collapse

Short answer

Model collapse is the degradation that occurs when AI models are trained heavily on AI-generated content rather than human-created data. Over successive generations, outputs grow less diverse, less accurate, and more prone to repeating errors. It is one reason original, human-produced content and high-quality data sources remain valuable as AI-generated material spreads online.