AI accusation against Las Vegas attorney raises ethical question in courts

AI accusation against Las Vegas attorney raises ethical question in courts

LAS VEGAS (KSNV) — A Las Vegas defense attorney is accused of using Artificial Intelligence (AI) in a court brief submitted in a case, after the prosecution claims it contains a citation to a nonexistent Supreme Court case.

According to the Chief Deputy District Attorney, prosecutors became suspicious after noticing Dean Kajioka's reply to the defense differed slightly from the original petition.

Prosecutors say Kajioka cited the opinion multiple times and quoted it as if it were legitimate.

According to a motion filed by the County DA's office, the prosecution is asking a judge to strike down a 70-page reply brief submitted by Kajioka and impose sanctions the court deems appropriate.

News 3 has reached out to Kajioka for comment. We are waiting for a response.

It has led to the question of whether AI should be used to defend someone in court.

"The thing about Artificial Intelligence is that it's designed as a default to try to please the user; it wants to make the user happy," said Matthew Hoffman, Managing Partner of Battle Born Injury Lawyers.

Hoffman says AI can have its limits.

"And if you don't give it limitations, it will-it's notorious for just making things up," said he said.

AI is becoming more common in the courtroom.

But Hoffman says you have to be careful about how you use it.

"It has access to a lot of information, but how it interprets it, it's not human. It will make mistakes, and if you rely on it, you will get in trouble," said Hoffman.

UNLV Law Professor Nancy Rapoport says nationally there are more than 1,600 cases so far where attorneys have been caught using "hallucinated materials."

Hallucinated materials, in simple terms, mean fabricated, fictional, or made-up case laws or legal references generated by AI.

"It absolutely can jeopardize the case because one of the things lawyers are supposed to do is make sure that what they file is truthful," said Rapoport. "So if you file something that is manifestly untrue, the court can throw out the pleading."

She says in some situations the court can also dismiss the case.

Rapoport and another UNLV law professor are trying to prepare students to properly use AI by teaching a new course to first-year law students.

"It's because lawyers have been making so many mistakes about how to use AI appropriately that we thought students have already used AI as undergraduates. We want to train them in how to use it ethically and responsibly," she said.

How will the class use AI as a tool?

"Students can use it to do first drafts of things if their law professor lets them," said Rapoport. "The main thing is to make sure that anything you hand to another person or to the court gets proofread and checked first."

She says the best way to proofread is to go back and reread the statutes and cases.

Overall, the argument is that how you use AI for legal purposes comes down to a moral code.

"So, if an attorney theoretically was to utilize AI to ask it to come up with stuff, and then you don't go check it, and it's all false or fake or wrong, and you sign off on it, that's a big ethical concern," said Hoffman.