I agree the lawyer shouldn't have trusted ChatGPT, but I'm not comfortable with the idea that the lawyer bears all the responsibility for using ChatGPT and Microsoft/OpenAI bear no responsibility for creating it.

"May occasionally generate incorrect information" is not a sufficient warning. Even Lexis-Nexis has a similar warning: "The accuracy, completeness, adequacy or currency of the Content is not warranted or guaranteed."

And in any case, it seems like you agree with me that the lawyer was incompetent rather than malicious.