I saw a New York Times article last
week about a lawyer for a plaintiff in a personal injury suit
against an airline who used CHATGPT to draft a brief opposing a
motion to dismiss his case. As it turned out, that was not a great
decision.
According to the Times article, the brief cited a litany of
supporting cases that were, um, made up. Lawyers for the airline
apparently searched unsuccessfully for the cases and finally let
the court know what they found. Or, more appropriately I guess,
what they didn't find.
The plaintiff's lawyer (I'm not going to name him
because there but for the grace of God go any number of us)
apparently threw himself on the mercy of the court, which was his
only option given that the only checking he did on the bogus cites
was to ask the CHATGPT program if they were accurate. The program
assured him they were. But it now appears AI is not the most
reliable source. The Judge set a hearing for June 8 to discuss
potential sanctions. So, no
