Universal Music Group and other music publishers, including Concord and ABKCO1, have filed a lawsuit against Anthropic, an AI company. The lawsuit alleges that Claude 2, Anthropic's large language model (LLM) and rival to OpenAI's ChatGPT, was trained on their copyrighted content.

The plaintiffs claim that Anthropic infringed on their copyrights, first by training its chatbot on protected songs without permission and then by illegally distributing the lyrics in its prompted answers, removing copyright management information in breach of the Copyright Act of 1976. The lawsuit demands a permanent injunction banning any copyrighted works' infringement and compensation for damages and attorney's fees.

Why does it matter?

The music industry has been trying to find ways to use the technology while still protecting its rights. The recent lawsuit could set a precedent for song lyrics and AI copyright. It could potentially compel AI companies to remove copyrighted works from the training datasets and the output generated by their LLM-powered chatbots.
Copyright infringement has become a controversial issue in generative AI. Legal challenges have been mounting recently for AI tools, and this lawsuit echoes comparable legal action from writers and actors suing AI firms over the use of their work. The decisions courts will reach in cases like this one may become the foundation for future rules overseeing AI.