Artificial Intelligence (AI) has become unavoidable and increasingly essential in a variety of fields, including the legal profession. While it offers many benefits when used responsibly, there are significant risks when attorneys fail to check the work product generated by AI tools. There have been numerous examples of attorneys submitting unverified AI content to courts, complete with cites to non-existent cases entirely created (or hallucinated) by AI. A recent decision of the Supreme Court, New York County Commercial Division, provides an excellent example of the negative consequences of this conduct to both the attorney and the client.

Ader v. Ader involved a family dispute, wherein the now deceased father agreed to provide a guaranty on a loan made to his son. The son agreed to repay his father if the father was, in fact, required to make any payments under the guaranty. After the father’s death, the son defaulted on the loan and failed to make the necessary repayments. While the facts are interesting in their own right, recent litigation in the case concerned the conduct of the son’s attorney.

The plaintiff moved for summary judgment, and when the defendant filed its brief in opposition to the motion, the plaintiff’s counsel identified a series of inaccurate citations and quotations in the defendant’s brief. In deciding for the plaintiff, the Court noted that it would not accept “fake quotations.” This was true even if they “happened to be arguably correct statements of law,” or the use of fake citations and quotations generally supported “real” statements of law. The Court expressed concern that because of the broad use of unvetted AI, a fake quotation or citation would eventually make its way into a Court’s decision.

This is a fairly common example of the misuse of AI in practice. However, there is still more in this case. The plaintiff’s counsel moved for sanctions against the defendant for his conduct. In opposing the motion for sanctions, which sought to punish him for using unvetted AI, the defendant’s counsel yet again submitted opposition papers which included more fake citations and quotations than in the first set of papers, including cites to four non-existent cases.

The defendant’s counsel was unable to adequately justify his conduct (or express appropriate remorse). As a result, the Court imposed monetary sanctions on both the attorney and his client. The Court summarized the problem with the conduct of the defendant’s counsel, stating: “When attorneys fail to check their work—whether AI-generated or not—they prejudice their clients and do a disservice to the Court and the profession. In sum, counsel’s duty of candor to the Court cannot be delegated to a software program.”

AI is a valuable tool, and it has quickly become an essential asset to every attorney’s practice. However, it is important to choose counsel who will use it as a tool to be leveraged, rather than a shortcut.

If you have a litigation matter, contact one of our attorneys to discuss how we can help.