A US judge has fined two lawyers and their law firm $5,000 (£3,935) after they submitted fake court citations generated by the artificial intelligence chatbot ChatGPT in a legal filing. The incident occurred in an aviation injury claim against Colombian airline Avianca.
District Judge P Kevin Castel in Manhattan ordered Steven Schwartz, Peter LoDuca, and their firm Levidow, Levidow & Oberman to pay the penalty. Schwartz admitted that ChatGPT invented six legal cases he included in a legal brief. The judge stated that while using AI for legal assistance is not inherently improper, lawyers must ensure the accuracy of their filings.
In his written opinion, Judge Castel said the lawyers 'abandoned their responsibilities when they submitted nonexistent judicial opinions with fake quotes and citations created by the artificial intelligence tool ChatGPT, then continued to stand by the fake opinions after judicial orders called their existence into question.' He noted that one fake decision had traits consistent with actual rulings but also contained 'gibberish' and was 'nonsensical'.
Levidow, Levidow & Oberman said in a statement that they 'respectfully' disagreed with the court's finding of bad faith, calling it a 'good-faith mistake in failing to believe that a piece of technology could be making up cases out of whole cloth.' Schwartz declined to comment, and LoDuca did not immediately respond.
ChatGPT had suggested several aviation mishap cases that Schwartz could not find through usual methods. Some cases were not real, misidentified judges, or involved non-existent airlines. Chatbots like ChatGPT, developed by OpenAI, can suffer from 'hallucinations' or inaccuracies, as seen in other instances where the AI falsely accused a law professor of sexual harassment or gave incorrect answers about the James Webb space telescope.
In a separate ruling, the judge dismissed the underlying aviation claim because the statute of limitations had expired.
