MIT Finds "Attribution Decay" in AI-Generated Art

MIT CSAIL researchers identified "attribution decay": as training datasets grow, any single image's influence on a model's output shrinks, making it hard to trace AI art back to specific training data.

2026-08-19 ~ 2026-08-19 · 2 related posts

• MIT study finds "attribution decay": removing any single artist's data often changes generated images not at all — MIT_CSAIL · 2026-08-19
• MIT study: AI art often can't be traced to training data — nordicinst · 2026-08-19