Following AI company Anthropic's settlement in a copyright infringement lawsuit, a conflict has emerged over the distribution of the settlement funds. While literary agents and publishers are demanding a share of the payout, authors are showing strong resistance.
This case stems from a legal dispute over the unauthorized use of copyrighted books as training data for Anthropic's generative AI models. While the settlement was seen as a crucial precedent regarding compensation for the use of copyrighted works in AI development, its distribution method has become a new subject of debate.
Amid the rapid development of generative AI, ongoing debates persist regarding rights management in LLM training sets and who should receive the resulting revenue. This conflict highlights the complex structure of copyright management in the digital age—specifically, to whom the revenue from intellectual property like books should belong: publishers, agents, or authors.
Groups of authors argue that fair compensation should be paid directly to the creators, who are the primary source, without intermediaries. This dispute is attracting significant attention as it could have a major impact on future contracting models between other AI companies and creators.