Mata v. Avianca: The Case That Changed AI in Law

On June 22, 2023, Judge P. Kevin Castel of the Southern District of New York issued an order that would be cited in hundreds of subsequent rulings, bar association guidelines, and legal ethics opinions. The order sanctioned attorneys Steven Schwartz and Peter LoDuca of Levidow, Levidow & Oberman for filing a brief containing six entirely fabricated judicial decisions generated by ChatGPT.

The underlying case, Mata v. Avianca, Inc., was a personal injury claim against an airline. The legal issues were routine. What made it extraordinary was the discovery that the brief supporting the opposition to dismissal cited cases that did not exist in any legal database.

When opposing counsel and the court could not locate the cited authorities, Schwartz submitted an affidavit explaining that he had used ChatGPT to supplement his legal research and that the AI had provided the citations. He stated that he had never used the tool before and was unaware that it could generate fabricated content.

The court found that the attorneys had acted in bad faith by failing to verify the citations before filing and by initially doubling down when the court raised questions about the authorities. The $5,000 sanction per attorney was accompanied by a requirement to notify every judge falsely cited in the fabricated opinions.

The immediate impact was judicial. Within six months, federal judges across the country issued standing orders requiring AI disclosure in filings. The Northern District of Texas, the Eastern District of Pennsylvania, and the District of Colorado were among the first. By mid-2024, the trend had spread to state courts.

The broader impact was cultural. Mata v. Avianca made every attorney aware that AI tools could produce convincing-looking citations that are entirely fake. It did not create a new obligation. It highlighted an existing one: attorneys are responsible for the accuracy of their filings, regardless of the tools they use to prepare them.

For the legal technology industry, the case created a category. Citation verification tools went from a niche idea to a recognized need. The question shifted from "why would I verify my citations?" to "how do I verify them efficiently?"