A lawyer used ChatGPT for legal filing. The chatbot cited nonexistent cases it just made up

Lawyer #StevenSchwartz of Levidow, Levidow & Oberman has been practicing law for three decades. Now, one case can derail his entire career.

He relied on #ChatGPT in his legal filings and the chatbot completely manufactured previous cases, which Schwartz cited

#LevidowLevidowOberman #Avianca #MataVAvianca #legal #law #artificialintelligence #ai #generativeAI #technology #tech

https://mashable.com/article/chatgpt-lawyer-made-up-cases

A lawyer used ChatGPT for a legal filing. The chatbot cited nonexistent cases it just made up.

The lawyer now may face sanctions for submitting the bogus cases.

Mashable
@gtbarry Honestly, I feel a little bit sorry for the guy. I mean, right, huge screwup and he richly deserves the humongous benchslap he's most assuredly going to get. Still, though: if he'd done the research the way we all think he should have, he'd have done it by sitting at a computer and looking things up in a piece of software called something like Lexis or Westlaw. He almost certainly thought that he was doing the same thing with a piece of software with a different name and a way easier user interface. That's absolutely wrong, of course, but not an unreasonable thing to think if you've sort of vaguely heard about this whole AI thing but been too busy to follow it in detail. He was probably in a hurry because this was a penny-ante PI case and he was on contingency, so he was basically working for free. Not an excuse, I'm not saying that, but it's understandable. I've been on deadline on cases that weren't paying me very much, and, you know, there but for fortune and all that.

@jonberger ChatGPT to save research time shouldn't be an issue. LLMs are social aggregation tools after all.

Lack of validating the results is my issue. It shows either a lack of discipline or competence