[toomuchtodo]

When your GPT is wrong, you will be laughed out of the room and sanctioned.

When your attorney is wrong, you get to point at the attorney and show a good faith effort was made.

Hacks are fun, just keep in mind the domain you're operating in.

[giantg2]

"When your attorney is wrong, you get to point at the attorney and show a good faith effort was made."

And possibly sue their insurance to correct their mistakes.

[Scoundreller]

But you’ll have to find a lawyer that specializes in suing lawyers and their own malpractice plans.

Maybe that’s where legal AI will find the most demand.

[kulikalov]

Can't a tech firm running a "legal gpt" have an insurance?

[giantg2]

No. Malpractice insurance would be at the professional level. There could be lawyers using a legal chatGTP, but the professional liabilities would still be with the licensed professional.

[kulikalov]

Well, I guess since it's not "practice" we gonna call it "mal-inference insurance".

[freejazz]

More legal malpractice? No, because they aren't attorneys and you cannot rely upon them for legal advice such that they'd be liable to you for providing subpar legal advice.

[kulikalov]

Why? Because there's no word for "insurance of AI advise accuracy"? The whole point of progress is that we create something that is not a thing at the moment.

[freejazz]

No, because, like I said, GPTs are not legally allowed to represent individuals, so they cannot obtain malpractice insurance. You can make up an entirely ancillary kind of insurance. It does not change the fact that GPTs are not legally allowed to represent clients, so they cannot be liable to clients for legal advice. Seeing as how you think GPTs are so useful here... why are you asking me these questions when a GPT should be perfectly capable of providing you with the policy considerations that underline attorney licensing procedures.

[giantg2]

That was the point of my comment - no ability to collect the insurance.

[toomuchtodo]

Do they have a license to practice law?