ABSTRACT The intersection of technology and the legal profession has evolved significantly, with legal practitioners using tools like Westlaw, LexisNexis, and Google Search for legal research. More recently, artificial intelligence (AI) tools, such as ChatGPT, have been integrated into legal practice, offering both promise and challenges. In particular, incidents like the Mata v. Avianca case, where attorneys faced sanctions for submitting fictitious citations generated by ChatGPT, highlight the risks of relying on AI in legal work. While courts have issued guidelines for the responsible use of AI, there remains a clear need for due diligence and a thorough understanding of its limitations. This study evaluates the effectiveness of AI tools in legal research by presenting two complex legal questions, one involving U.S. asylum law and the other concerning Colombian legal principles. The responses from ChatGPT and Google's AI were analyzed for accuracy and consistency, revealing significant gaps in both tools' performance, particularly in providing complete and current legal information. While they can assist in research, they are not yet reliable enough to replace expert legal analysis. The findings suggest that legal professionals should approach AI‐generated legal information with caution and verify results with human expertise.
David Yeregui Marcos del Blanco (Wed,) studied this question.