Anthropic Faces Scrutiny Over AI Citation Error

A federal judge has ordered Anthropic to respond to allegations that it submitted a court filing containing a false citation generated by its Claude AI chatbot in a copyright case brought by Universal Music Group, Concord, and ABKCO. The dispute involves Anthropic's alleged use of copyrighted song lyrics to train its AI. Anthropic's expert witness, Olivia Chen, cited a non-existent academic article, and while the publishers' attorney initially suggested the citation was a fabrication, Anthropic's lawyers later clarified that Claude had generated incorrect citation details rather than fabricating an entire article. They apologized for the error and acknowledged that their manual review failed to detect the mistake. The judge emphasized the seriousness of AI-generated citation errors and hallucinations, reflecting growing court concerns about AI-assisted legal work. The case highlights broader legal issues over the use of copyrighted material in generative AI training.