A new study from MIT researchers indicates that image-generating AI models such as Midjourney and Stable Diffusion produce outputs that are impossible to attribute to specific training data. This "attribution decay" means that as training datasets grow, tracing the origin of any particular generated image becomes increasingly difficult. The findings present a significant hurdle for artists attempting to pursue copyright infringement claims against AI companies. AI
IMPACT Challenges copyright claims against AI companies by making it impossible to prove if specific artistic works were used in training.
RANK_REASON Research paper from MIT on AI model attribution. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Mastodon — fosstodon.org →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →