Report from the Bloomberg
In Brief – A New Jersey US district court judge withdrew his decision in a biopharma securities case after lawyers pointed out that his opinion contained numerous errors, including made-up quotes and misstated case outcomes. Judge Julien Neals completely withdrew his denial of the motion to dismiss a shareholder lawsuit made by CorMedix Inc. and posted that a new opinion and order will be forthcoming. Earlier in the month, lawyers in a separate securities case in the same district pointed out flaws in Neals’ CorMedix opinion, saying it “contains pervasive and material inaccuracies” and “includes quotations from case law and pleadings that cannot be found in the sources cited.” Although none of the formal filings with the court allege that Judge Neals used an AI chatbot for writing his opinion, the flaws in the document, including fabricated quotes, mis-referenced cases, and completely incorrect interpretations of decisions, carry the hallmarks of AI “hallucinations” that have come to the fore in legal matters as lawyers increasingly rely on AI services and chatbots to prepare case documents.
Context – Chatbot “hallucinations” in legal briefs first made news in mid-2023, less than a year after ChatGPT’s release, with two lawyers being sanctioned in US District Court. Chief Justice John Roberts led off his 2023 Annual Report on the Federal Judiciary discussing AI technology in the legal system, warned about hallucinations, and urged responsible use. It seemed primarily aimed at the bar. Well, similar cases keep popping up, with lawyers facing various sanctions. Of course, everyone should know that they need to check AI work, but everyone knows they should not drink and drive, too. The Judicial Office of the UK Ministry of Justice released AI guidance in late 2023 to judges and courts, including about the use of chatbots in official court duties. They warned of the limitations of the systems, including “hallucinations”, as well as the fact that most AI training materials at the time was US-centric and might not reflect UK legal tradition. Everyone, please always check your AI chatbot work. And please don’t drink and drive.