Another day, another misuse of ChatGPT. A Colorado attorney was fired from his job after using, and suspended last month for one year and one day (with all but 90 days stayed, subject to probation) because he used, ChatGPT to prepare a motion. As with other lawyers who’ve gotten into trouble for misusing AI, Zachariah Crabill filed the motion without verifying that case citations were accurate. Lo and behold,  they were not.

 According to the Colorado Supreme Court decision, Crabill discovered the cases were fictitious (the decision does not elaborate), and did not do what he should have done and withdraw the pleading or at least alert the Court to the problem. When the judge inquired, he “falsely attributed the mistakes to a legal intern.” Days after the motion hearing, Crabill filed an affidavit explaining that he used ChatGPT.

The parties stipulated to four violations of the Colorado Rules of Professional Conduct (which largely track the Model Rules)—1.1/competency, 1.3/diligence, 3.3(a)(1)/candor to the tribunal, and 8.4(c)/dishonesty. These all make sense, and probably account for the suspension—the cover-up, here, is worse than the “crime.” 

Part of the “cover-up”—blaming the intern—particularly bugs me. In addition to the dishonesty, it’s rude, and also? If the lie were true, it still wouldn’t help, and would possibly implicate Rule 5.1. Boss types—the buck stops with you.