SALT LAKE CITY (ABC4) A Utah lawyer has been sanctioned by the Utah Court of Appeals after a filing he made was found to use ChatGPT and contain a reference to a nonexistent court case. The brief was written by a law clerk at the firm, and the false citations were not reviewed before filing.

The Utah Court of Appeals has decided to impose sanctions on Richard Bednar, an attorney who they claim filed a brief that used false citations created by ChatGPT.

According to documents, Bednar and Douglas Durbano, as the petitioner’s counsel, filed a “timely petition for interlocutory appeal.” When the respondent’s counsel reviewed this brief, they found it to reference cases that were incorrect or did not exist.

“It appears that at least some portions of the Petition may be AI-generated, including citations and even quotations to at least one case that does not appear to exist in any legal database (and could only be found in ChatGPT and references to cases that are wholly unrelated to the referenced subject matter,” the respondent’s counsel wrote, according to documents.

The referenced case was “Royer v. Nelson,” a case that did not exist in any legal database and was found to be an “AI ‘hallucinated’ case that does not exist anywhere other than on ChatGPT.”

In a phone call with Matthew C. Barneck, the attorney who represented Bednar at the hearing, he shared that Bednar was not aware that ChatGPT was used until the citations were called into question. At that point, Bednar owned up to what happened.

Following this response, Bednar “acknowledged ‘the errors contained in the petition’ and apologized,” according to the document from the Utah Court of Appeals. He also offered to pay attorney fees incurred by the response to “make amends.”

At a Utah Court of Appeals hearing on April 22, 2025, documents say that Bednar and his attorney for the hearing “acknowledged that the Petition contained fabricated legal authority, which was obtained from ChatGPT, and they accepted responsibility for the contents of the Petition.”

They told the court that an “unlicensed law clerk” had prepared the brief, and Bednar did not “independently check the accuracy” before he signed it and filed it with the court. Durbano was reportedly uninvolved in the creation and filing of the petition. The law firm did not have an AI policy in place at the time, but they told the court on April 22 that they had since initiated one.

Barneck clarified that most law clerks are not licensed, and the law clerk in question had previously been licensed and was a law school graduate.

According to the analysis by the Utah Court of Appeals, this is the first time a court filing has been created with AI, and it has not been previously addressed by the Utah State Courts. They turned to other cases to make their decision on the sanctions.

“We agree that the use of AI in the preparation of pleadings is a legal research tool that will continue to evolve with advances in technology. However, we emphasize that every attorney has an ongoing duty to review and ensure the accuracy of their court filings. In the present case, Petitioner’s counsel fell short of their gatekeeping responsibilities as members of the Utah State Bar when they submitted a petition that contained fake precedent generated by ChatGPT,” the Utah Court of Appeals said in the document.

The Utah Court of Appeals further stated that, while they appreciated Bednar’s acknowledgment of his wrongdoing, they still needed to place sanctions due to the mishandling of the filing and not thoroughly reviewing the document before he signed it.

Bednar was ordered to pay the respondent’s attorney fees for the petition and hearing, refund fees to their client for the time used to prepare the AI filing and attend the hearing, and donate $1,000 to “and Justice for all” — a Utah legal nonprofit — within 14 days.

“Although the Office of Professional Conduct (OPC) is bound by strict confidentiality rules and cannot comment on pending or nonpublic matters, the OPC takes seriously any conduct that may compromise the integrity of the judiciary and the legal profession. Additionally, the Utah State Bar is actively engaging with practitioners and ethics experts to provide guidance and continuing legal education on the ethical use of AI in law practice,” the Utah State Bar said in a statement to local affiliate ABC4.

On the ethics section of their site, the Utah State Bar discusses their thoughts on the use of AI. They offer the following advice to lawyers who are considering the use of ChatGPT: “Treat AI-generated results like a draft from a law clerk. Check any propositions or citations that you’re not 100% sure about.”

Barneck told ABC4 that what happened was an oversight, and the court’s opinion indicates that they believe it was done negligently and not recklessly. He shared that he was glad that Bednar owned up to what happened instead of trying to sweep it under the rug like lawyers in a similar position have.