Judge Brantley Starr of the U.S. District Court for the Northern District of Texas has issued a standing order requiring attorneys to include in all filings a certificate attesting either that (a) no portion of the filing was drafted by artificial intelligence (i.e., ChatGPT), or (b) that any language drafted by artificial intelligence was checked for accuracy by a human being. Judge Starr observed that while artificial intelligence platforms are “incredibly powerful and have many uses in the law,” they should not be relied upon for legal briefing because they are “prone to hallucination and bias.” For instance, they are known to “make stuff up” like case citations and quotations. They are also unreliable or biased because they “hold no allegiance to any client, the rule of law, or the laws and Constitution of the United States (or…the truth).”
This order follows an incident in New York where a lawyer is facing potential sanctions for relying on ChatGPT to draft a legal brief which included citations to fake cases.