Too Far With ChatGPT

Two New York lawyers filed a brief with “fake law” in it, having accepted at face value a “hallucination” by ChatGPT abouot several cases that did not in fact exist. The district court has now sanctioned them. The opinion is of broad general interest, given the great influence of generative AI in its few short months of widespread public availability. And it’s also of broad general interest in avoiding sanctions–the lawyers had at least one good chance, and maybe even two, of coming clean about their error, but instead “doubled down” with purported screenshots of the nonexistent cases.