PR Supreme Court issues warning about AI use after lawyers submitted filing with false citations.

Puerto Rico Supreme Court Justice Raúl Candelario López

By THE STAR STAFF

The Supreme Court of Puerto Rico has dismissed an ethics complaint against two attorneys but issued a formal warning after they submitted a court filing rife with false case citations, which the lawyers said were generated with the help of artificial intelligence (AI).

The lawyers had already been fined $1,000 by the Court of First Instance, which the Supreme Court called the first known monetary penalty in Puerto Rico tied to improper AI use in litigation. In its opinion, issued as 2026 TSPR 41, the court said a repeat could lead to suspension from the practice of law.

Justice Raúl Candelario López, writing for the court, said the justices did not impose additional discipline because of “mitigating circumstances,” including that the $1,000 fine had been imposed and promptly paid, the attorneys had withdrawn from the underlying case and apologized, and they had no prior disciplinary history. Still, the court stressed that lawyers must verify any material produced by technological tools and act with competence and diligence -- an obligation underscored in the current Rules of Professional Conduct, including Rule 1.19’s duty of technological competence.

The dispute began when the lawyers filed a brief citing nonexistent or inaccurately described precedents. Opposing counsel flagged the citations and asked for sanctions. The attorneys responded that the errors stemmed from the challenges of using new technology and said they never meant to mislead the court, adding that they regretted the mistakes.

The court analyzed the conduct under Puerto Rico’s former Code of Professional Ethics, citing Canons 18 (competent, diligent work), 35 (honesty and truthfulness), and 38 (protecting the dignity of the profession). It also cited Rule 9.1 of the Rules of Civil Procedure, which requires lawyers to make a reasonable inquiry to ensure filings are grounded in fact and supported by law.

Those duties, the justices said, are basic: before presenting authorities to a court, lawyers must check that the cases exist and stand for what the filing claims. Submitting invented or wrongly attributed citations reflects deficient legal research and a failure to meet minimum professional standards, the court said, whether the source is a traditional database, a general internet search, or an AI tool.

After concluding that the conduct violated ethical duties, the court weighed common discipline factors -- such as harm, remorse, whether the episode was isolated, and the lawyers’ prior record -- and chose a warning rather than another sanction. The opinion said that similar future misconduct could result in suspension.

The justices said the current professional rules are designed to confront emerging technologies, including AI. Rule 1.19, the court noted, requires lawyers to maintain a reasonable understanding of technological developments and to use tools with knowledge of their benefits and risks -- meaning AI-assisted drafting must be carefully reviewed for accuracy, relevance, and legal correctness, especially before anything is filed in court.

In a concurring opinion, Associate Justice Camille Rivera Pérez -- joined by Associate Justice Luis Estrella Martínez -- agreed with dismissing the complaint and issuing a warning but emphasized that repeat conduct could merit harsher discipline. The concurrence said court proceedings are a search for truth and noted that disputes over AI-driven errors are surfacing in other jurisdictions as the technology spreads through everyday legal practice.