Judge suspects Uprise attorneys used AI with phony cases, quotes in filing (www.rgj.com)

A Washoe County judge has ordered Uprise’s attorneys to appear in court after finding what he says are multiple fictitious case citations and made‑up quotes in their filings — errors he strongly suspects were produced by generative AI. District Judge David Hardy flagged that “at least fourteen of the decisions Uprise cites appear to be fictitious,” and that other real references include invented language or reversed conclusions. The dispute arises in a suit over unpaid work on a Nevada fiber‑optic project tied to allegations that Uprise’s CEO, Stephen Kromer, misappropriated $7 million. The named lawyers, Jan Tomasik and Daniel Mann of Cozen O’Connor, must explain the sourcing of their authorities at a Sept. 5 hearing and face potential sanctions, Bar referral, or removal from the case. The episode is a concrete, high‑stakes example of “hallucinated” AI output leaking into legal practice: generative models can fabricate plausible but nonexistent precedents and quotations, creating risks that courts and litigants will waste time, reach incorrect holdings, or lend legitimacy to fake authorities. Beyond immediate sanctions, the incident underscores technical and ethical implications for AI-assisted legal drafting — the need for rigorous human verification, provenance tracking of sources, and professional responsibility rules to evolve as lawyers adopt large language models.