Nippon Life v. OpenAI tests UPL liability for ChatGPT filings

The US District Court for the Northern District of Illinois is weighing whether OpenAI can be held liable for the unauthorized practice of law when a non-lawyer uses ChatGPT to draft court filings. In *Nippon Life Insurance Co. of America v. OpenAI Foundation*, the insurer alleges that Graciela Dela Torre, after settling a disability claim with her employer's insurance company in January 2024, uploaded her correspondence to ChatGPT, asked it whether she had been 'gaslit,' fired her attorneys, and then produced 'two lawsuits with over 60 filings and a subpoena, all written by ChatGPT,' including fabricated citations.

OpenAI's defense, as laid out at TechPolicy.Press by University of Chicago Law School legal assistant Daniel Kalish, runs in two directions at once. The company calls ChatGPT 'a helpful tool… that advances access to justice,' and in the same filings argues it 'does not and cannot have the requisite level of knowledge and intent' to aid and abet UPL. Nippon Life answers that 'under traditional products liability doctrine, [OpenAI] is responsible for that tool,' and that model training and chat surveillance show OpenAI 'manifested the intent to interfere' with the settlement.

Two Illinois UPL statutes are in play: the Attorney Act, dating to the mid-1800s, and the Corporation Practice of Law Prohibition Act of 1917. Kalish writes that AI-flagged writing has turned up in over 18 percent of pro se litigant filings, and has grown annually since ChatGPT's release. First Amendment and antitrust defenses are flagged in the piece as possible next moves, not yet decided.