Lawsuits allege AI companies used copyrighted content to train models without permission
Musicians, artists, and news publishers are suing AI companies including OpenAI and Stability AI, alleging that copyrighted content was used to train generative AI models without permission or compensation. Courts have begun ordering discovery, including the production of user logs and the evaluation of copyright infringement claims across multiple lawsuits.
What happened
Universal Music Group, Warner Records and Sony Music Entertainment sued AI music startups, including Udio, for allegedly using their recordings to train text-to-music models without permission.
The New York Times and other newspapers sued OpenAI over use of copyrighted content to train its generative AI model. In November 2025, a federal judge ordered OpenAI to produce 20 million anonymized ChatGPT user conversations to the plaintiffs.
In the class-action case Andersen v Stability AI, artists allege that use of their artwork to train AI image generators without credit, compensation or consent violates their rights.
Reported impact
A January 2024 survey found that more than a third of illustrators had lost income due to AI, and one study projects a 21% revenue loss for audiovisual creators by 2028.
Artists, publishers and studios argue that the use of their material in AI training is copyright infringement and that creative workers bear the brunt of the AI boom, with generative AI already displacing creative labour.
Index analysis
Dozens of lawsuits are proceeding across US courts as judges struggle to reconcile copyright law with technology that undermines the concept of authorship.
UMG subsequently announced a deal with defendant Udio to create an AI music platform, but the Music Artists Coalition responded that artists remain sidelined despite industry assurances of partnership.
Governance controls
The Human Artistry Campaign, an industry-artist coalition, has rallied creatives and executives to jointly endorse legislation protecting artists from AI and big tech.
The court-ordered discovery process includes protective orders removing identifying information from user logs handed over by OpenAI.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.