"18 Lawyers Caught Using AI Explain Why They Did It"

404 Media (Jason Koebler & Jules Roscoe) has the details.

Comments

[Michael Ejercito]

They wanted to save labor costs from having to hire paralegals?

[Bruce Hayden]

Law is a business, and that means that there are always financial pressures to cut corners. My view is that it is on the courts and lawyer discipline organizations to make the costs of screwing up high enough that these things are not tolerated. One suggestion might be for judges to toss briefs after maybe the 2nd fictitious cite - similar to the rule by some judges to cut off briefs that exceed page limits.

What should be easy to accomplish is a cite checker that makes sure that cited cases actually exist. What is more difficult is one that determines that cases cited for a proposition actually stand for that proposition. Probably every attorney here has run into that situation pre-AI. It’s no doubt worse now with AI. Sometimes, it’s blatant, and those times tools should ultimately be able to detect them. But sometimes it is subtle, and it’s going to be harder.

And also, I would love to see a tool that could detect when court holdings were no longer applicable. A couple years ago, I ran into a long string cite here at Volokh for a proposition. It even included a Supreme Court case. Except that almost the entire string was invalid because Congress had amended the relevant statute in response to the Supreme Court decision, and every cited case, except for one (from a different Circuit), was decided based on the old law. That sort of thing is going to be much harder to detect.

[Stupid Government Tricks]

And also, I would love to see a tool that could detect when court holdings were no longer applicable.

Here's a simple idea: get rid of precedent. Make all decisions from first principles, starting with the Constitution and the law in question. Stop pretending that what courts decided 100 years ago is relevant.

ETA: Of course this will never happen. It would make law libraries and case histories and all that research and billable hours unnecessary. It would allow ordinary people to understand laws, even if they didn't have the experience to go toe-to-toe with a lawyer who argues cases every day.

[David Nieporent]

For every complex problem there is an answer that is clear, simple, and wrong. - HL Mencken

[Stupid Government Tricks]

Case in point: you, quoting simple quotes in defense of complex billable hours.

[Malika la Maize]

Simple quotes are fine for meeting simple problems and people. They’re not good for complex things, which the world is full of

[Malika la Maize]

If you fancy yourself a conservative you might want to read a little less Brietbart and a little more (or some) of Burke of Kirk.

[Stupid Government Tricks]

If you fancy yourself a giraffe ...

Oh who am I kidding. You're just a goof responding to voices inside your head.

[Darth Buckeye]

When I use AI I tell it what facts to use, what precedents to use, and what conclusions to reach. I use it only for composition. I limit using it for research; I learned that most precedents it cites are non-existent.

Sometimes I ask it a question about a legal issue. I treat any answer the way I would treat an article in an old legal encyclopedia: with interest, but with a large amount of scepticism.

[dwb68]

I find this one the most interesting:

"The lawyer “acknowledge[d] the cited authorities were inaccurate and mistakenly verified using Westlaw Precision, an AI-assisted research tool, rather than Westlaw’s standalone legal database.” The lawyer further wrote that she “now understands that Westlaw Precision incorporates AI-assisted research, which can generate fictitious legal authority if not independently verified. She testified she was unable to provide the Court with this research history because the lawyer who produced the AI-generated citations is currently suspended from the practice of law in Louisiana:"

Later, a similar story about Lexis.

I can certainly see the confusion there.