DOJ Lawyer Resigns Over AI-Generated Court Brief
A U.S. Justice Department attorney resigned after a federal judge discovered that a court filing contained fabricated quotes and inaccurate legal citations that appear to have been generated using artificial intelligence. The incident has raised serious concerns about the growing use of AI tools in legal work and the responsibility of attorneys to verify the accuracy of documents submitted to courts.
The case involved an assistant U.S. attorney in the Eastern District of North Carolina who had filed a legal brief in a federal lawsuit involving the U.S. Department of Defense. The lawsuit itself was brought by a retired Air Force colonel challenging a policy affecting coverage of GLP-1 weight-loss medications under the TRICARE for Life health program for military retirees.
During the litigation process, the plaintiff—who is also a lawyer—noticed that several quotations and legal citations included in the government’s filing did not appear to match the cases cited. After reviewing the document, the plaintiff alerted the court to what he believed were serious inaccuracies in the brief.
U.S. Magistrate Judge Robert Numbers reviewed the filing and confirmed that multiple problems existed in the document. According to the judge’s order, the brief contained fabricated quotes from court decisions, misstatements about the holdings of certain cases, and even quotations that appeared to come from federal regulations but could not be verified.
The judge expressed “serious concerns about the accuracy” of the statements and citations included in the filing. Because legal briefs are expected to accurately represent case law and regulatory authority, the discovery raised the possibility of sanctions against the attorney responsible for submitting the document.
Initially, the assistant U.S. attorney explained that the problematic citations had resulted from filing an unfinished draft document by mistake. According to the explanation referenced in court filings, incorrect citations were included inadvertently when the document had not yet been finalized.
However, the explanation did not resolve the court’s concerns. Judge Numbers ordered a hearing requiring senior leaders from the U.S. Attorney’s Office to appear and explain why the attorney should not be sanctioned and why the entire office should not share responsibility for the errors.
Possible sanctions discussed by the court included financial penalties, contempt proceedings, or even suspension from practicing before the court. The judge also directed the head of the U.S. Attorney’s Office for the Eastern District of North Carolina to review the issue and determine what corrective action should be taken.
Before the scheduled hearing could take place, the Justice Department attorney involved submitted a resignation. The resignation effectively prevented the court from formally reprimanding the lawyer in open proceedings, though the matter still highlights the risks associated with improper use of generative AI in legal practice.
Legal experts say the incident reflects a growing problem in the legal profession. As generative AI tools become more widely used, courts have seen an increasing number of filings containing “hallucinated” citations—references to cases or legal language that do not actually exist. In some instances, lawyers have relied too heavily on AI-generated text without independently verifying the information before submitting it to courts.
Judges across the United States have begun warning attorneys that they remain fully responsible for everything filed in court, regardless of whether the text was drafted by humans or generated by AI systems. Courts expect lawyers to check every citation and quote carefully before submitting documents.
The situation also highlights the ethical obligations lawyers face under the Federal Rules of Civil Procedure. Attorneys must ensure that legal arguments and authorities cited in their filings are accurate and supported by law. Submitting fabricated citations—even unintentionally—can undermine the integrity of judicial proceedings.
The controversy comes as the legal profession continues to grapple with how to integrate artificial intelligence tools into everyday practice. Many law firms are experimenting with AI to assist with research, drafting, and document review, but courts and regulators have emphasized that human lawyers remain accountable for the final product.
While the resignation resolved the immediate issue of sanctions in this case, the incident is likely to fuel further discussions about AI governance in the legal field. Judges, bar associations, and law firms are increasingly exploring rules and policies designed to ensure that AI tools are used responsibly and that attorneys maintain professional oversight.
Ultimately, the case serves as a cautionary example for lawyers experimenting with AI technologies. Even powerful new tools cannot replace careful legal research and professional judgment, and mistakes made through automated systems can carry serious consequences in the courtroom.
Key Legal Outcomes
-
A Justice Department attorney submitted a legal brief containing fabricated quotes and incorrect citations.
-
A federal judge identified the inaccuracies and raised concerns about possible sanctions.
-
The court scheduled a show-cause hearing to determine whether disciplinary penalties were warranted.
-
Potential sanctions included fines, contempt proceedings, or suspension from court practice.
-
The attorney resigned before the judicial hearing could take place.
Why This Matters
-
Highlights the growing risks of AI-generated legal errors in court filings.
-
Demonstrates that lawyers remain legally responsible for verifying AI-produced content.
-
Shows courts are increasingly scrutinizing the use of generative AI in litigation.
-
Raises ethical questions about professional oversight when using AI drafting tools.
-
May influence future rules governing AI usage in the legal profession.
