Artificial Intelligence or AI seems to be permeating every facet of life from education to our court system. And when AI makes mistakes, as it frequently does, if the humans involved don’t check their facts it can cause big problems. I recently spoke to Senator Chuck Grassley, head of the Judiciary Committee, who’s taking on the issue as it impacts America’s courtrooms.

The following is a transcript of a report from “Full Measure with Sharyl Attkisson.”
Watch the video by clicking the link at the end of the page.

At issue is AI-generated information allegedly used by two federal judges who got caught issuing error-riddled court rulings: Judge Henry Wingate of Mississippi and Judge Julien Neals of New Jersey.

On July 20, Judge Wingate issued an order that paused a Mississippi law banning diversity, equity and inclusion or DEI agendas in public schools.

Sharyl: Judge Wingate in Mississippi, apparently, according to your letter, had serious inaccuracies in his order, including ‘naming plaintiffs and defendants that weren’t parties to the case, misquoting state law, factually inaccurate statements, and referencing four individuals who don’t even appear in the case.’

Senator Chuck Grassley: It’s so obvious that he should be embarrassed. And I hope that at the Supreme Court level, the Chief Justice is over the entire judicial branch of government, and he’s got an office called the Administrative Office of the Courts. I would hope that they would take disciplinary action and also set a standard.

After Judge Wingate’s flawed order, the party he’d ruled against— the Attorney General of Mississippi— pointed out the errors. Senator Grassley says Wingate then issued a backdated “corrected” version, removed the original order without explaining, dismissing the mistakes as “clerical.”

The same week Judge Neals in New Jersey withdrew his flawed decision in a case. He’d used inaccurate quotes, and misstated the record. For example, he wrote that motions to dismiss were denied when they’d been granted.

In response to questions from Grassley, both judges blamed low-level staffers who drafted their opinions and orders using Perplexity and Chat GPT. Both say they’ve implemented measures to prevent a repeat. And the Administrative Office of the U.S. Courts, which makes policy for federal courts, told Senator Grassley it’s working to develop permanent policies on use of AI.

Sharyl: Short of having some sort of federal standard, what do you say is the obligation of judges that may use, or their clerks may use, artificial intelligence for research?

Grassley: Their obligation is to question everything they read, but they should use it very sparingly, and obviously not include that information in your courts as we’ve seen so many errors. In other words, a decision of a judge should not have any errors.

Sharyl: Do you have any overarching concerns about artificial intelligence? I understand the applications on the Judiciary Committee of the judiciary, but this is a research tool as ubiquitous as Google are going to become that if it’s not already being used by everybody for all kinds of research.

Grassley: Well, you know, you read it mostly about college students, using it to write news, write their term papers and things of that nature. It might be the same thing for a college professor, for somebody that’s got students that are using AI to get information. But the whole job of a judge is to question all the facts of the case and very astutely read the law and apply it accordingly. You expect them to take the extra step just like you’d expect a college professor to take the extra step to make sure that somebody isn’t using AI to actually cheat. And in a sense, these judges are cheating their clients or the people that are before their court.

Sharyl (on-camera): Grassley says his office has been contacted by AI industry whistleblowers. He’s proposing a law to offer them job protection when they blow the whistle.

Watch video here.