# Can "My Clerk Used ChatGPT Without My Knowledge" Save a Utah Lawyer from Sanctions?

**Short answer: It very likely does not, though it may soften the punishment.** The defense runs into several structural features of Utah's ethics rules that place the ultimate responsibility on the lawyer, not on whoever (or whatever) generated the text.

## The Rules That Do the Heavy Lifting

- **Rule 3.3(a)(2) – Candor Toward the Tribunal.** A lawyer must not knowingly offer evidence or argument the lawyer *knows* to be false. The word "knows" is doing real work, but Utah (like most states) construes knowledge to include situations where the lawyer *should have known* through reasonable care. A lawyer who files a string of nonexistent cases without checking them fails that care standard regardless of whether the error came from a clerk, a vendor, or an LLM.

- **Rule 5.3 – Responsibilities Regarding Nonlawyer Assistance.** The lawyer who directs or supervises a nonlawyer (a law clerk falls here) must make *reasonable efforts* to ensure the nonlawyer's conduct is compatible with the lawyer's professional obligations. That includes instituting a citation-verification step. "I didn't realize the clerk was using ChatGPT" is essentially an admission that the reasonable-efforts obligation wasn't met.

- **Rule 5.1 – Supervisory Lawyers.** For partners and managing attorneys, the duty is explicit: take reasonable steps to ensure the work of subordinate lawyers and staff complies with the Rules.

- **Rule 8.4(c) – Conduct involving dishonesty, fraud, deceit, or misrepresentation.** Even without intent to deceive, presenting fabricated authority to the court is a misrepresentation of the record.

The through-line: **the lawyer is the gatekeeper.** The clerk (human or machine) is an instrument. The bar disciplines the lawyer because the lawyer's name is on the filing and the lawyer vouched to the court that the authorities exist.

## Why the "I Didn't Know" Explanation Weakens Rather Than Rescues

1. **It converts an honest mistake into a supervision failure.** A judge hearing "my clerk fabricated these" already has a problem. Adding "…and I had no idea the clerk was using an AI" signals the attorney lacked a review protocol. In a pre-AI fabrication case the same logic applied: "I didn't know the intern made them up" never excused the filing attorney.

2. **It may undermine a good-faith narrative.** A Utah disciplinary panel or a reviewing court may infer the lawyer was sufficiently detached from the work product to be unaware of the *method* of research, which raises the question of how much the lawyer actually reviewed the brief.

3. **It does not negate the underlying duty.** Even a perfectly supervised clerk who manually fabricates a citation produces the same sanctionable outcome. The source of the error is irrelevant to the court's integrity interest.

## What *Might* Matter as Mitigation

Utah's disciplinary process (and Rule 8.4 itself) weighs circumstances. A lawyer can argue:

- **No intent to deceive.** The citations were plausible-sounding AI hallucinations, not deliberate inventions.
- **Prompt, voluntary correction.** The lawyer disclosed the error to the court, opposed the sanctions, and moved to withdraw the offending pages before the court relied on them.
- **Reasonable reliance in an emerging area.** In 2023–2024, when AI-hallucination guidance was still sparse, a court may exercise some forbearance it wouldn't in 2026, when the Utah State Bar and the Utah Supreme Court have had time to issue explicit expectations.
- **Institutional safeguards now in place.** Demonstrating a post-incident policy (e.g., mandatory Westlaw/Lexis "verified" citation checks, a second-attorney review, a no-AI-without-supervision rule) can show the firm learned.

These factors can push the outcome from disbarment/suspension toward a reprimand, a modest fine, or a requirement of CLE in legal-ethics or legal-research methods. They do not, in the main, eliminate sanctions.

## Practical Posture If You're the Lawyer in Question

- **Disclose to the court immediately**, in writing, identifying the specific fabricated citations, how they were detected, and the steps taken to correct the record.
- **Do not over-explain the AI angle in the disciplinary filing.** A brief, factual statement ("The citations were generated by an AI tool used by staff and were not verified against the reporters before filing") is defensible. A long narrative about who clicked what button reads like an excuse.
- **Preserve the work-product trail.** Screenshots of the ChatGPT output, the clerk's notes, any review memos—these become evidence of whether reasonable verification steps existed.
- **Cooperate with the Utah State Bar's grievance process** and consider a conditional settlement (admit-and-agree) that includes a fine and CLE rather than a contested hearing.
- **Consult a legal-ethics specialist** in Utah (or the ABA's Center for Professional Responsibility) before responding to any notice of misconduct.

## Bottom Line

Utah's rules make the *filing attorney* the accountable party for the accuracy of everything presented to the tribunal. "My clerk used ChatGPT and I didn't know" addresses *how* the error occurred, not *whether* the lawyer breached the duty of verification and supervision. It may be relevant to the *degree* of sanction, but as a full defense it will almost certainly fail. The strongest protective strategy is swift, transparent correction to the court, documented good-faith efforts to remedy the record, and credible evidence that the firm now has verification controls in place.

---

*This is general legal-ethics information, not legal advice. Specific outcomes depend on the facts, the court, the tribunal, and evolving bar guidance. A Utah-licensed ethics attorney should be consulted on any live matter.*