US

Lawyer In New York Uses ChatGPT For Court Filing. It Does Not Go Well.

(Photo by LIONEL BONAVENTURE/AFP via Getty Images)

Dana Abizaid Contributor
Font Size:

A lawyer representing a man suing Avianca airlines for negligence admitted to briefly using ChatGPT to produce evidence that turned out to be fake, according to CNN.

Robert Mata sued Avianca for injuries allegedly sustained from an airline serving cart in 2019, according to the CNN report. Steven Schwartz, an experienced lawyer with New York law firm Levidow, Levidow and Oberman, represented Mata.

At least six cases that Schwartz submitted as research for a brief, however, “appear[ed] to be bogus judicial decisions with bogus quotes and bogus internal citations,” Judge Kevin Castel of the Southern District of New York said in a May 4 order, per CNN.

“The court is presented with an unprecedented circumstance,” the judge wrote, adding that the plaintiff’s counsel submitted a motion that “is replete with citations to non-existent cases.” (RELATED: Elon Musk Tells Tucker Carlson AI Could ‘Absolutely’ Take Control Of Civilization)

ChatGPT reportedly generated the sources in question.

Schwartz submitted the following suspect cases: Varghese v. China South Airlines, Martinez v. Delta Airlines, Shaboon v. EgyptAir, Petersen v. Iran Air, Miller v. United Airlines, and Estate of Durden v. KLM Royal Dutch Airlines. The judge and defense noted that none of these cases appeared to exist, the filing said, according to CNN.

Schwartz responded in an affidavit that this was the first time he had used ChatGPT as a research tool and, therefore, “was unaware of the possibility that its content could be false.”

Schwartz accepted responsibility for not confirming the sources and said in an affidavit this week that he “greatly regrets having utilized generative artificial intelligence to supplement the legal research performed herein and will never do so in the future without absolute verification of its authenticity.” He has since apologized.

Fellow attorney Peter Loduca said he “had no reason to doubt the sincerity” of Schwartz’s research.

In an affidavit filed on April 25 Schwartz was ordered to show why he shouldn’t be sanctioned “for the use of a false and fraudulent notarization.”

In his defense, Schwartz filed an affidavit Wednesday that reportedly showed ChatGPT confirming the authenticity of the cases.

When asked whether or not the Varghese case “is a real case,” the chatbot doubled down and claimed, “Upon double-checking, I found the case Varghese v. China Southern Airlines Co. Ltd., 925 F.3d 1339 (11th Cir. 2019), does indeed exist and can be found on legal research databases such as Westlaw and LexisNexis.”

The chatbot also said that the other cases “are real” and can be found on “reputable legal databases.”

On June 8, Schwartz will face a sanctions hearing over his use of the chatbot.