Written by: Olivia Salvador
On March 10, 2026, Assistant U.S. Attorney Rudy Renfer of North Carolina resigned after filing an AI-written brief to the court. At the time, Renfer was representing the Defense Department in an underlying lawsuit by a North Carolina litigant challenging a policy limiting availability of GLP-1 weight loss medications for TRICARE for Life participants. Renfer submitted a brief to the court which included fabricated quotes and misstated the holdings of several cases.
Upon receipt of the brief, Magistrate Judge Robert Numbers suspected AI had been used to prepare the writing. Noticing the lack of accuracy of certain quotes and positions in Renfer’s brief, Numbers held a hearing where he questioned and reprimanded Renfer for his “disappointing conduct.” Renfer claims that he “accidentally overwrote and lost a prior version of the filing, feeling panicked, he had AI rewrite it, then filed it thinking he’d reviewed it when he filed.” Although Renfer took responsibility for the serious mistake and acknowledged that he had spread his attention too thin while was working on multiple filings, this error ultimately cost him his job as an Assistant U.S. Attorney.
Numbers pointed out that the use of AI to take a short cut on otherwise “basic work” made this situation “all the more outrageous.” He also made it clear that improper use of AI in the legal field can seriously damage one’s career.
Background
Artificial Intelligence (AI) is technology that enables computers and machines to simulate and perform tasks that would typically require human intelligence. It can imitate human abilities like thinking, learning, understanding, and problem solving. AI uses algorithms to process large amounts of data and find patterns within it. Over time, AI improves its performance by learning from new data, which helps it give better responses and more accurate decisions.
AI can be grouped into three different categories: Narrow, General and Superintelligent AI.
Narrow AI refers to systems that are designed to carry out specific tasks or solve problems within a limited area. This includes your typical question and answer platforms. Popular examples of Narrow AI are Siri and Alexa. These AI systems can’t think or make decisions beyond what they are programmed to do. They work by understanding the question asked, analyzing the words and their meanings, and generating a response by predicting the most likely and helpful answer.
General AI refers to a type of AI that can understand, learn, and apply knowledge across different tasks. Its capabilities go beyond those of Narrow AI. Unlike Narrow AI, General AI can transfer what it has learned in one situation to another, solve complex problems, learn new skills on its own and adapt to unfamiliar situations. Currently, General AI does not exist in any platform. It is a theoretical concept subject to ongoing research.
Superintelligent AI is another theoretical type of AI that would surpass human intelligence in all areas. It would be capable of thinking, innovating and reasoning far beyond human ability. Superintelligent AI could learn and adapt to new situations at a “superhuman pace,” greatly exceeding the speed and capacity of the human brain.
Today, Narrow AI is the most used form of AI and can be found in many of our daily interactions, including voice assistants like Siri and Alexa, recommendation systems such as Netflix suggestions based on viewing history, self-driving or assisted driving features in vehicles and customer service chat boxes.
AI in the Legal Field
AI is increasingly becoming a popular tool used within the legal industry. AI tools can increase productivity and efficiency by expediting tasks that would normally take an associate more time to complete. Legal research platforms now use AI to help locate case law and other relevant legal information within their platform based on keywords entered by the user. These platforms can also analyze and summarize cases, assist with document drafting and review, and perform other legal research tasks. Additionally, law firms are using AI tools to draft memos, assist with discovery review, support litigation analysis, conduct contract review, perform legal research, and predict case outcomes.
Although AI can be helpful, it also has the potential to create “hallucinations.” AI “hallucinations” happen when legal AI systems produce “fabricated case citations, distorted holdings, or false procedural information that appears authentic but does not actually exist or is factually incorrect.” Given this possibility, best practices include carefully verifying AI generated outputs while prioritizing human judgment and expertise.
Conclusion
The Renfer incident highlights a critical lesson: while AI can be a powerful and helpful tool, it can not replace good lawyering. AI is not a substitution for human judgment and responsibility. Renfer’s mistake in trusting AI’s output without carefully reviewing for its accuracy cost him his role as a DOJ Attorney. This case serves as a reminder that AI should be used as a support, not a shortcut, and that taking the time to review AI-generated work is essential to maintaining professionalism.
Sources:
A legal practitioner’s guide to AI & hallucinations, TRI/NCSC AI Policy Consortium for L. & Courts, (Feb. 16, 2026).
AI Ethics Lab, Rutgers Univ., (last visited Mar. 27, 2026).
Artificial intelligence: What it is, how it works and why it matters, ISO, (last visited Mar. 27, 2026).
Chris Williams, DOJ Attorney Throws Himself Under the Bus Rather than Dragging Down Everyone Else, Above The Law, (Mar. 11, 2026).
Cole Stryker & Eda Kavlakoglu, What is artificial intelligence (AI)?, IBM, (last visited Mar. 27, 2026).
Generative Artificial Intelligence, U.S. Mil. Acad. Libr., (last updated Dec. 30, 2025).
How Is AI Impacting the Legal Profession?, Vand. Univ. L. Sch., (Jan. 20, 2026).
Kyle Jahner, DOJ Lawyer Quits Before Judicial Scolding for AI Brief Error (1), Bloomberg Law, (Mar. 10, 2026).
Robert J. Couture, The Impact of Artificial Intelligence on Law Firms’ Business Models, (Feb. 24, 2025).
Types of AI: Explore Key Categories and Uses, Syracuse Univ., (Mar. 26, 2025).