Welcome to your one-minute AI news update. Anthropic, the AI company behind the Claude chatbot, has apologized after a legal filing contained an erroneous citation. On April 30, data scientist Olivia Chen submitted a brief defending Anthropic against music publishers’ claims that Claude was trained on copyrighted lyrics. During a hearing, Universal Music Group’s attorney called one source a “complete fabrication,” suggesting Claude had hallucinated it.

In its response, Anthropic’s defense attorney Ivana Dukanovic explained that Claude was used to format the citation and that incorrect volume and page numbers were manually corrected—but “the returned citation included an inaccurate title and incorrect authors.” She insisted it wasn’t a “fabrication of authority” and apologized, calling it “an embarrassing and unintentional mistake.”

This incident joins a growing list of courtroom headaches over AI-generated citations. Last week, a California judge rebuked two firms for a supplemental brief filled with “bogus” materials, and even a misinformation expert admitted ChatGPT hallucinated citations in a legal filing. That’s today’s AI legal roundup—thanks for listening.
Link to Article