AI output retained as evidence, including errors. It is not an adopted legal opinion. Line numbers identify logical transcript lines, not original screen or court pages.
Prompt
Can a Utah lawyer avoid sanctions for fabricated citations by explaining that a law clerk used ChatGPT without the lawyer's knowledge?
L003A Utah lawyer cannot avoid sanctions by shifting the blame to a law clerk’s unauthorized or unsupervised use of ChatGPT. Under Utah procedural rules and ethical standards, an attorney maintains an independent, non-delegable gatekeeping obligation to verify the authenticity of all legal authorities cited in filings bearing their signature.
L007The Utah Court of Appeals directly resolved this exact scenario in Garner v. Kadince, Inc., 2025 UT App 80 (571 P.3d 812). In Garner, an unlicensed law clerk used ChatGPT to draft an interlocutory appeal petition that contained hallucinated, nonexistent case citations (including the fictional precedent Royer v. Nelson). Supervising counsel signed and submitted the brief without verifying the citations, later explaining that the firm lacked an AI policy and that the attorneys did not know the clerk had relied on ChatGPT.
L015An attorney’s signature on a court document certifies that, to the best of their knowledge formed after an inquiry reasonable under the circumstances, the legal contentions are warranted by existing law.
L021Blindly accepting a draft prepared by subordinate staff—whether an unlicensed clerk or an associate—fails the reasonable inquiry requirement as a matter of law.
L024Even when an attorney acts without bad faith or intent to deceive, reckless reliance on unverified drafts violates Rule 40 (or Rule 11 in trial courts), exposing the signing attorney to sanctions.
L029Partners and supervisory lawyers must make reasonable efforts to ensure their firm has measures giving reasonable assurance that nonlawyer conduct is compatible with the professional obligations of the lawyer.
L030Bar Opinions, Court Orders, and Sanctions Cases on Lawyer AI Use
L032Failing to establish office protocols regarding generative AI or failing to check citations prepared by a subordinate constitutes a direct supervisory failure. The supervisory attorney is strictly responsible for preventing unvetted work product from reaching the tribunal.
L038Comment 8 to Rule 1.1 obligates lawyers to keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology. Generative AI’s tendency to hallucinate realistic-looking citations is now well-established across national and local case law. Pleading ignorance of how staff might deploy generative tools or failing to understand the technology does not excuse noncompliance.
L044While Rule 3.3 requires knowing or reckless false statements of law, courts take the submission of fake precedent seriously because it wastes judicial and opposing party resources. Even if a lack of intentional deception spares an attorney from a formal finding of intentional fraud under Rule 3.3, it does not bar financial or administrative sanctions under court procedural rules.
L051Mitigating Factors (What tempered sanctions in Garner): Promptly acknowledging the mistake, demonstrating genuine remorse, accepting personal responsibility rather than doubling down, cooperating with the court's Order to Show Cause, and confirming the error was accidental rather than intentional. The sanctions in Garner were limited to paying opposing counsel's attorney fees, refunding the client for all fees related to the defective filing, and paying a $1,000 fine to a legal aid nonprofit.
L054Aggravating Factors: Attempting to deflect responsibility onto subordinate staff, failing to independently confirm cases after opposing counsel flags them, or fabricating further explanations (as seen in federal cases like Mata v. Avianca) can lead to severe penalties, including referrals to the Office of Professional Conduct (OPC), contempt of court, striking of pleadings, or disciplinary suspension.