Lawyers Cited for Using Fake Cases from ChatGPT Must Compensate

Three attorneys and their law firm have been ordered to pay sanctions after being caught citing completely fabricated cases in legal filings, using an AI language model to make up quotes and citations. OpenAI’s ChatGPT language model was used to invent made-up judiciary rulings which Peter LoDuca, Steven A. Schwartz and the firm of Levidow, Levidow & Oberman P.C then submitted to a judge. After speaking to the AI model, Schwartz himself then lied when confronted about the legitimacy of the cases he cited. The judge’s sanctions order included a $5,000 fine for each attorney and an acknowledgement to each real judge falsely identified as the author of the cited fake cases. The judge also dismissed the underlying injury claim in the original case because it was filed too long after the incident. This incident highlights the importance of lawyers taking responsibility for what they file with the court, even when using tools that may help enhance their output, according to legal industry sources.