AI Slip-Up: Legal Turmoil as Anthropic’s Chatbot Faces Allegations in Mysterious Courtroom Blunder

A lawyer representing Anthropic was compelled to issue an apology after using an erroneous legal citation generated by the company’s Claude AI chatbot in a federal court filing. The incident took place amid an ongoing lawsuit between Anthropic and music publishers over accusations of copyright infringement.

According to court documents filed on Thursday in a Northern California court, Claude mistakenly produced a reference that contained both an incorrect title and fictitious authors. Anthropic’s legal team explained in the filing that their routine manual review process failed to spot the fabricated citation, along with several other inaccuracies also produced by Claude.

Anthropic characterized the incident as an “honest citation mistake,” explicitly denying any intentional fabrication or misrepresentation of legal authority.

The error drew immediate scrutiny after opposing attorneys for Universal Music Group and other publishers accused Anthropic’s expert witness, company employee Olivia Chen, of relying on AI-generated citations that included non-existent articles. Following this allegation, federal judge Susan van Keulen ordered Anthropic to address these charges directly.

This case highlights another instance of complications arising from lawyers improperly relying on artificial intelligence in legal contexts. Earlier this same week, a California judge strongly criticized two law firms for presenting AI-generated misinformation as legitimate research. In a separate incident in January, an Australian lawyer similarly faced backlash after court filings prepared with the assistance of ChatGPT included citations of cases that did not exist.

Despite these high-profile missteps, investments aimed at introducing AI into the practice and facilitation of legal work continue to grow. One prominent legal tech startup, Harvey, which builds generative AI tools for lawyers, is reportedly seeking funding exceeding $250 million at an estimated valuation of $5 billion.

More From Author

“Revolutionary Breakthrough: The Invisible World of Microbial Sensors That Spy on Hidden Threats”

Are Microsoft’s Coders Becoming Obsolete? The AI Shift Behind Mysterious Layoffs

Leave a Reply

Your email address will not be published. Required fields are marked *