Celeste Greaves
Volume 77, Issue 3, 659-688
Artificial intelligence (“AI”) burst onto the legal scene and will soon revolutionize practice altogether. Despite enduring technical issues, like hallucinations and implicit bias, AI’s capabilities and potential benefits have created massive impetus toward adoption. Recent cases like Mata v. Avianca highlight the growing need for adaptation in our existing frameworks to ensure the continued legitimacy of representations to the court. The “reasonable inquiry” standard of Rule 11 of the Federal Rules of Civil Procedure is woefully unequipped to meet this challenge for three primary reasons. First, traditional methods for determining whether a new technology has a high level of efficacy, and can therefore be relied upon, do not apply to AI. Second, when human interactions with AI create negative outcomes or mistakes, it disrupts conventional methods for assigning blame. And third, the “black box” inherent in generative models precludes the retrospective analysis and reasonable inquiry required to analyze the legitimacy of legal reasoning. Finally, even if a reasonable inquiry were possible, there is no legal consensus defining what constitutes “reasonable” in the context of AI. Ultimately, amending Rule 11 is necessary to better tailor the rule to new challenges faced by the judiciary in a rapidly evolving technological landscape.