The Case for Grace

In case you missed it, a plaintiff’s lawyer in New York was recently featured in the New York Times for submitting a document to a court that was drafted by ChatGPT and replete with fabricated case law resulting from artificial intelligence hallucinations.[i]  Plaintiff’s counsel has since become a subject of national ridicule and undoubtedly violated numerous rules of professional conduct (e.g., the duties of competence, confidentiality, and supervision to name a few),[ii] for which he—and his firm—are facing sanctions.[iii]

The issue came to light in defense counsel’s response to plaintiff’s filing, indicating that defense counsel was unable to find the cases cited in plaintiff’s filing.  This, of course, prompted the judge to request copies of the case law.  Plaintiff’s counsel submitted alleged copies of the decisions, which were again fabrications drafted by artificial intelligence, apparently unbeknownst to plaintiff’s counsel.[iv] Defense counsel then wrote a letter to the judge, questioning the authenticity of the cases identified.

Defense counsel and others in his firm suspected “the cases in the brief were not real . . . [and] had an inkling a chatbot might have been involved.”[v]  In that moment, defense counsel faced a question many attorneys encounter:  in the face of an obvious error by opposing counsel, what do you do?

Here, defense counsel chose to alert the court, a choice that was certainly available under the model rules of professional conduct.  But this was not his only option, as “[t]he Rules do not . . . exhaust the moral and ethical considerations that should inform a lawyer.”[vi]  Defense counsel could have contacted plaintiff’s counsel first to address the issue.  By all accounts, it appears that plaintiff’s attorney legitimately did not know or understand the inherent risks associated with answers provided by ChatGPT.[vii]   (Whether he should have known is a different question, and the answer to that question is undoubtedly yes.)[viii]  But there’s no reason to believe that plaintiff’s counsel was acting with any malicious or deceptive intent when filing the document.[ix]  And, had defense counsel contacted plaintiff’s counsel first, he could have alerted plaintiff’s counsel to the fabricated case law and likely discovered the source of the fabrications.  Defense counsel could have shared his own knowledge of ChatGPT’s limitations regarding legal research, which would have allowed plaintiff’s counsel the opportunity to withdraw or amend the filing without the need for judicial involvement. And all of that could have been accomplished without risking the defense’s position in the matter.

I do not mean to suggest that defense counsel did anything unethical or immoral.  My point, instead, is that defense counsel—who admittedly recognized the potential source of the issue—could have given plaintiff’s counsel—who obviously failed to understand the limitations of a new technology—a bit of grace in the situation. Grace is defined in many ways, including “an act or instance of kindness, courtesy, or clemency,”[x] “favor or goodwill,”[xi] and “a disposition to kindness and compassion.”[xii] Giving grace would have helped not only plaintiff’s counsel to save face but also the legal profession to maintain credibility as a self-regulating entity.[xiii]  Many judges have advised that, “[w]hile serving as advocates for their clients, lawyers are not required to abandon notions of civility.”[xiv]  For, “[i]f the bar is to maintain the respect of the community, lawyers must be willing to act out of a spirit of cooperation and civility and not wholly out of a sense of blind and unbridled advocacy.”[xv]    

Generative AI is likely to change our profession ultimately for the better, but we are in the midst of a learning curve. Accordingly, we should seize these teachable moments and show grace to our fellow bar members while we all learn how to use—and not use—this new tool.

[i] Benjamin Weiser, Here’s What Happens When Your Lawyer Uses ChatGPT, New York Times (May 27, 2023), available at https://www.nytimes.com/2023/05/27/nyregion/avianca-airline-lawsuit-chatgpt.html.

[ii] See Karen Sloan, A lawyer used ChatGPT to cite bogus cases. What are the ethics?, Reuters (May 30, 2023), available at https://www.reuters.com/legal/transactional/lawyer-used-chatgpt-cite-bogus-cases-what-are-ethics-2023-05-30/.

[iii] Sara Merken, Lawyer who cited cases concocted by AI asks judge to spare sanctions, Reuters (June 8, 2023), available at https://www.reuters.com/legal/transactional/lawyer-who-cited-cases-concocted-by-ai-asks-judge-spare-sanctions-2023-06-08/.

[iv] Plaintiff’s counsel asked ChatGPT if the cases it cited were real cases, and it confirmed that they were.  Plaintiff’s counsel asked for the source of that information, and ChatGPT replied that the cases could “be found in reputable legal databases.”  Weiser, supra note i.

[v] Weiser, supra note i.

[vi] 2023 ABA Model Rules of Professional Conduct, Scope ¶ 16, available at https://www.americanbar.org/groups/professional_responsibility/publications/model_rules_of_professional_conduct/model_rules_of_professional_conduct_preamble_scope/.

[vii] Wes Davis, A lawyer used ChatGPT and now has to answer for its ‘bogus’ citations, The Verge (May 27, 2023), available at https://www.theverge.com/2023/5/27/23739913/chatgpt-ai-lawsuit-avianca-airlines-chatbot-research. According to affidavits filed in the case, plaintiff’s counsel relied primarily on Fastcase for legal research in other state-law matters and did not have access to federal cases, so he turned to ChatGPT, which he misunderstood to be a search engine, and ChatGPT falsely advised plaintiff’s counsel that the cases it provided could be found on both Westlaw and LexisNexis.

[viii] “A lawyer must have a reasonable basis for believing any statement that he makes to a tribunal.” Douglas R. Richmond, The Ethics of Zealous Advocacy: Civility, Candor and Parlor Tricks, 34 Tex. Tech L. Rev. 3, 29 (2002).

[ix] When seeking the court’s mercy at the sanctions hearing, plaintiff’s counsel noted that he has “’suffered professionally and personally’ and is embarrassed and humiliated.” Merken, supra note iii.

[x] https://www.merriam-webster.com/dictionary/grace

[xi] https://www.dictionary.com/browse/grace

[xii] https://www.vocabulary.com/dictionary/grace

[xiii] The title of the New York Times article is “Here’s What Happens When Your Lawyer Uses ChatGPT” (emphasis added), suggesting that all attorneys are likely to misuse generative AI and repeat the blunder made by plaintiff’s counsel. Had defense counsel reached out to plaintiff’s counsel before involving the court, media involvement would likely have been greatly diminished or even nonexistent, thus avoiding the negative implications for the profession as a whole.  It is our responsibility, as members of the bar, to “further the public’s . . . confidence in the rule of law and the justice system because legal institutions in a constitutional democracy depend on popular participation and support to maintain their authority.” 2023 ABA Model Rules of Professional Conduct, Preamble ¶ 6, supra note vi.

[xiv] Butts v. State, 273 Ga. 760, 772 (2001) (Benham, C.J., concurring).

[xv] Evanoff v. Evanoff, 262 Ga. 303, 304 (1992).