New York lawyers use artificial intelligence chatbots.
A New York lawyer faced a judge's rebuke after citing non-existent cases that were found to have conducted legal research using the AI chatbot ChatGPT. In one case involving an airline company being sued for personal injury, the plaintiff's counsel submitted a brief statement listing several cases as legal precedents. Unfortunately, it was later admitted in an affidavit that the following cases were found to be non-existent.
According to the BBC, the study was compiled by a veteran lawyer named StevenA. He allegedly has more than 30 years of experience practicing Chat, saying in the affidavit that he had not previously used ChatGPT for legal research and was unaware that its contents could be false. The lawyer asked if the chatbot Varghese was a real case? The chatbot replied yes.
When asked to provide a source, it told lawyers that the case could be found in legal research databases such as Westlaw and LexisNexis. When asked if the other cases you provided are fake? When it replied no. He added that the cases can be found in the same database. Chatbots may be interesting or seem very advanced, but they are still prone to hallucinations and answer answers that sound very coherent but have nothing to do with the real world.
Without rigorous fact-checking, using chatbots to research illusory legal cases that rely on real-world precedent rather than cool autocomplete isn't actually a tool to use.
The lawyer wrote that he deeply regrets the use of generative AI to supplement the legal research carried out this time, and vowed never to do so without absolute verification of its authenticity. JavaScript and Peter will face a hearing on June 8 to hear the incident.
