People v. Crabill is the case that showed what happens when an attorney doesn't just file AI-fabricated citations but lies about it afterward. In November 2023, Colorado attorney Zachariah Crabill received a suspension of one year and one day, with ninety days actively served, for using ChatGPT to draft a motion with fake cases and then blaming a legal intern when the judge questioned them.
Background
Zachariah Crabill, a Colorado attorney, used ChatGPT to draft a motion to set aside a judgment. The motion contained citations to cases that didn't exist, generated wholesale by the AI tool.
Before a scheduled hearing on the motion, Crabill discovered that the citations were fictitious. He knew the cases were fake before he walked into the courtroom. Despite this knowledge, he did nothing. He didn't alert the court, didn't withdraw the motion, and didn't file a corrected brief.
The case reached the Colorado Supreme Court's Office of Presiding Disciplinary Judge, which handles attorney discipline in the state. The proceedings revealed a pattern of conduct that went well beyond a single AI mistake.
What Happened
At the hearing, the judge questioned Crabill about the citations that couldn't be located. Rather than admitting he used ChatGPT, Crabill blamed a legal intern. He told the court that the intern had drafted the motion and was responsible for the fabricated authorities.
This was a lie. No intern had worked on the motion. Crabill had generated the entire document using ChatGPT and filed it under his own name. When the lie unraveled, Crabill eventually filed an affidavit with the court admitting that he had used ChatGPT to create the motion and that the citations were AI-generated fabrications.
The disciplinary proceedings identified multiple violations: Crabill failed to provide competent representation, failed to act with diligence, made false statements to the tribunal, and engaged in conduct involving dishonesty. The cover-up transformed a correctable AI mistake into a multi-count disciplinary case.
The Ruling
The Presiding Disciplinary Judge imposed a suspension of one year and one day, with ninety days actively served and the remainder stayed upon successful completion of a two-year probation period. Crabill violated multiple Colorado Rules of Professional Conduct.
The suspension was structured as a stipulated discipline, meaning Crabill agreed to the terms rather than fighting the charges. The one-year-and-one-day threshold is significant in Colorado because it requires the attorney to formally petition for reinstatement, unlike shorter suspensions where reinstatement can be more automatic.
The ruling treated the cover-up as seriously as the original fabrication. Filing fake citations was bad. Knowing they were fake before the hearing and doing nothing was worse. Lying to the judge's face about an intern was the conduct that pushed the case from sanctions territory into suspension territory.
Outcome: Crabill received a one-year-and-one-day suspension, with ninety days actively served and the remainder stayed upon successful completion of a two-year probation period. He violated multiple Colorado Rules of Professional Conduct.
Why This Case Matters
People v. Crabill was one of the first cases to result in an actual license suspension for AI misuse. While Mata v. Avianca produced a $5,000 fine, Crabill produced a suspension that prevented the attorney from practicing law. The escalation sent a clear message: the legal profession treats AI fabrication seriously, and dishonesty about it exponentially increases the consequences.
The case established a critical principle for attorney discipline: the cover-up is worse than the crime. Courts across the country now cite Crabill for the proposition that an attorney who discovers AI-generated fabrications has an immediate duty to disclose and correct, not conceal. Every day of silence after discovery compounds the violation.
Crabill also highlighted the candor-to-tribunal obligation. Rule 3.3 in most jurisdictions requires attorneys to be honest with the court. Fabricating an excuse (blaming a nonexistent intern) is an independent violation that carries its own severe consequences, regardless of what caused the original error.
Lessons for Attorneys
The most important lesson from Crabill isn't about AI. It's about what to do when you realize you've made a mistake. The moment Crabill discovered the citations were fake, before the hearing, his only viable path was immediate disclosure. Call the court, file a notice of correction, withdraw the motion. Courts consistently treat prompt self-correction as a mitigating factor.
Never blame someone else for your AI mistake. Crabill's decision to pin the fabrication on a legal intern turned a disciplinary issue into a dishonesty charge. Dishonesty to the tribunal is one of the most severely punished violations in legal ethics. It goes to the core of an attorney's fitness to practice and is far harder to recover from than a competence lapse.
Build a pre-filing verification checkpoint into your workflow. Before any motion or brief goes out, a second attorney or a trained paralegal should independently verify every citation. This creates a safety net that catches AI hallucinations before they reach the court and creates a documented chain of review that demonstrates diligence if something slips through.
The Bottom Line
People v. Crabill proved that the cover-up is worse than the AI mistake. An attorney who discovers fabricated citations and stays silent, or worse, lies about it, faces suspension rather than a fine. Immediate disclosure is the only defensible path.
AI-Assisted Research. This piece was researched and written with AI assistance, reviewed and edited by Manu Ayala. For deeper takes and the perspective behind the research, follow me on LinkedIn or email me directly.