Editor's note: The IAPP is policy neutral. We publish contributed and analysis pieces to enable our members to hear a broad spectrum of views in our domains.
In recent months, some Argentine courts have issued rulings on legal briefs in which the attorneys involved included citations of case law that, upon judicial review, were determined to be false or inaccurate, most likely generated using AI. Previously the subject of debate primarily in U.S. courts, Argentine courts are now beginning to assess the scope of lawyers' professional responsibility when employing AI technologies.
This discussion took place in three cases in different jurisdictions: the Court of Appeals in Civil, Commercial, Family, and Mining Matters of General Roca; Chamber II of the Civil and Commercial Court of Rosario; and Chamber I of the Civil and Commercial Court of Morón.
In all cases, the courts noted the submissions cited nonexistent decisions or judgments inconsistent with the transcribed text. In some cases, the precedents could not be found in any case law databases. For instance, in "Giacomino, César Adrián et al. v. Monserrat, Facundo Damián et al.," the Rosario Court of Appeals ordered a procedural measure to verify the citations and requested an explanation from the attorney, who expressly admitted having used an AI tool to prepare and reinforce his arguments. On the other hand, in "M.J.L. v. Peugeot Citroën Argentina S.A. et al.,"the court issued a similar measure, but the various parties did not clarify whether they used AI tools to prepare the submissions in question. Finally, in "Acevedo, Gerardo Gabriel v. Cáceres Mareco, Willian Arsenio," although no explanations were formally required from counsel, the court concluded, based on the drafting style of the filings, that AI tools had been used.
The judges agreed that the briefs had been prepared with the support of AI tools, which itself is not objectionable, but without adequate professional oversight. This sparked a debate on the scope of professional liability in the use of such tools. The courts recalled that, even when acting in good faith, submitting briefs that cite nonexistent case law undermines essential principles of the profession, including honesty, loyalty, truthfulness and good faith, all of which are provided for in the codes of ethics of the various Argentine jurisdictions.
In the specific cases under review, the courts decided not to impose direct sanctions on the lawyers. Nonetheless, they deemed it appropriate to notify the local bar associations, both to raise awareness of the risks and liabilities involved in using AI and to promote a broader discussion on responsible AI use in legal practice.
It was emphasized in all rulings that the issue is not the use of AI per se, whose potential as a supporting tool is undeniable, but rather using it without the subsequent verification that professional responsibility demands. They also emphasized that it is widely known that these technologies often "hallucinate" — that is, generate information that appears to be real but is not — cite sources out of context, or produce inaccurate references. As these risks are already documented and warned about in academic and practical forums, the level of diligence expected of lawyers increases proportionally.
Within this framework, the courts were very clear in stating that technology cannot replace the intellectual work of counsel nor absolve them of the duty to verify the accuracy of all information included in their briefs. The foreseeable possibility of error obliges attorneys, more than ever, to reinforce truthfulness and relevance checks. This entails corroborating each quote, retaining documentary backup and providing hyperlinks to facilitate judicial verification.
These precedents, along with others issued in different jurisdictions worldwide, are of interest not only in Argentina but also internationally. They demonstrate that the impact of AI is no longer a hypothetical scenario or one confined to the most technologically advanced jurisdictions. At the same time, they highlight that while AI is a valuable tool, its use requires human oversight to ensure the accuracy of submissions.
DiegoFernández is a partner at the Argentine firm Marval O'Farrell Mairal and is a member of the IAPP's research advisory board.
