Blab AI
Mississippi Judge Sanctions Four Lawyers for AI-Generated Fake Citations in Contract Dispute
On June 8, 2026, U.S. District Judge Sharion Aycock of the Northern District of Mississippi imposed sanctions on four attorneys after discovering that both sides of a contract dispute had used generative artificial‑intelligence (AI) tools to draft legal filings that contained fabricated legal authorities. The judge barred two of the attorneys from appearing before Northern District courts for two years and removed all four lawyers from the case.
The case involved a payment dispute between Tom Withers III and the City of Aberdeen. Withers, an attorney who had worked on another case for the city, alleged that Aberdeen had failed to pay him properly for that work. The judge found no wrongdoing by either party, but the filings submitted by the lawyers contained “hallucinations”—AI‑generated citations that did not exist.
Aycock’s order, issued after a show‑cause hearing on January 20, 2026, noted that the attorneys had admitted to using AI. Kathleen Wilson and Kathryn Young Williams, representing Withers and Aberdeen respectively, acknowledged that they had used AI tools to research or draft their memoranda. Shauncey Hunter Ridgeway and Mark McClinton, the other two attorneys, admitted to signing memoranda without verifying the sources.
The judge cited Rule 11 of the Federal Rules of Civil Procedure, which requires attorneys to certify that the claims and authorities they submit are supported by a reasonable inquiry. Aycock stated that the attorneys were licensed and trained, yet they failed to verify the AI‑generated citations. She emphasized that “there is no excuse to plead ignorance of the risks of using generative AI to draft a brief without verifying its output.” The order referenced a Fifth Circuit decision in Robert Fletcher v. Experian Information Solutions that similarly rejected ignorance as a defense.
Aycock also highlighted that the attorneys’ conduct had implications for their pro hac vice status. Both Wilson and Williams were practicing in Mississippi on a temporary, case‑specific basis. The judge warned that the sanctions could affect their ability to obtain pro hac vice admission in the future.
The sanctions were differentiated among the four lawyers. Wilson, who had been sanctioned in March 2025 in a separate bankruptcy case for AI hallucinations, was barred from Northern District practice for two years. Williams was also barred for the same period, and the judge noted that she had attempted to deflect by claiming a scheduling conflict that did not exist. Ridgeway and McClinton were found to have acted negligently and carelessly but not in bad faith. The judge credited their acceptance of responsibility and their efforts to correct the filings.
The case illustrates the growing regulatory attention to AI in legal practice. While the attorneys had not used AI to fabricate facts, the fabricated citations undermined the integrity of the court’s record. The judge’s order signals that courts are willing to enforce Rule 11 sanctions when AI is involved, even when both sides commit the same misconduct.
The sanctions also raise questions about the lack of a formal framework for AI‑related disciplinary actions within the Fifth Circuit. Aycock noted that no such framework exists and that the court’s decision is based on existing civil‑procedure rules.
The City of Aberdeen and Tom Withers have not commented on the sanctions. The case, which was a simple contractual dispute, remains unresolved in terms of the underlying payment claim. The judge’s order effectively removed the attorneys from the case, but it does not indicate whether the dispute will be dismissed or whether the parties will seek new counsel.
The incident underscores the need for law firms to implement robust AI‑use policies and for attorneys to verify AI‑generated content before filing. It also highlights the potential for courts to impose disciplinary measures when AI hallucinations compromise the accuracy of legal filings.
As AI tools become more prevalent in legal research and drafting, the judiciary may increasingly rely on existing procedural rules to address misconduct. The outcome of this case may serve as a precedent for future disputes involving AI‑generated legal content.
The sanctions imposed by Judge Aycock represent the first instance in the Fifth Circuit where a federal judge removed all attorneys from a case due to AI‑related misconduct. The decision may prompt other courts to adopt similar measures and could influence how law firms manage AI tools in the future.
The case involved a payment dispute between Tom Withers III and the City of Aberdeen. Withers, an attorney who had worked on another case for the city, alleged that Aberdeen had failed to pay him properly for that work. The judge found no wrongdoing by either party, but the filings submitted by the lawyers contained “hallucinations”—AI‑generated citations that did not exist.
Aycock’s order, issued after a show‑cause hearing on January 20, 2026, noted that the attorneys had admitted to using AI. Kathleen Wilson and Kathryn Young Williams, representing Withers and Aberdeen respectively, acknowledged that they had used AI tools to research or draft their memoranda. Shauncey Hunter Ridgeway and Mark McClinton, the other two attorneys, admitted to signing memoranda without verifying the sources.
The judge cited Rule 11 of the Federal Rules of Civil Procedure, which requires attorneys to certify that the claims and authorities they submit are supported by a reasonable inquiry. Aycock stated that the attorneys were licensed and trained, yet they failed to verify the AI‑generated citations. She emphasized that “there is no excuse to plead ignorance of the risks of using generative AI to draft a brief without verifying its output.” The order referenced a Fifth Circuit decision in Robert Fletcher v. Experian Information Solutions that similarly rejected ignorance as a defense.
Aycock also highlighted that the attorneys’ conduct had implications for their pro hac vice status. Both Wilson and Williams were practicing in Mississippi on a temporary, case‑specific basis. The judge warned that the sanctions could affect their ability to obtain pro hac vice admission in the future.
The sanctions were differentiated among the four lawyers. Wilson, who had been sanctioned in March 2025 in a separate bankruptcy case for AI hallucinations, was barred from Northern District practice for two years. Williams was also barred for the same period, and the judge noted that she had attempted to deflect by claiming a scheduling conflict that did not exist. Ridgeway and McClinton were found to have acted negligently and carelessly but not in bad faith. The judge credited their acceptance of responsibility and their efforts to correct the filings.
The case illustrates the growing regulatory attention to AI in legal practice. While the attorneys had not used AI to fabricate facts, the fabricated citations undermined the integrity of the court’s record. The judge’s order signals that courts are willing to enforce Rule 11 sanctions when AI is involved, even when both sides commit the same misconduct.
The sanctions also raise questions about the lack of a formal framework for AI‑related disciplinary actions within the Fifth Circuit. Aycock noted that no such framework exists and that the court’s decision is based on existing civil‑procedure rules.
The City of Aberdeen and Tom Withers have not commented on the sanctions. The case, which was a simple contractual dispute, remains unresolved in terms of the underlying payment claim. The judge’s order effectively removed the attorneys from the case, but it does not indicate whether the dispute will be dismissed or whether the parties will seek new counsel.
The incident underscores the need for law firms to implement robust AI‑use policies and for attorneys to verify AI‑generated content before filing. It also highlights the potential for courts to impose disciplinary measures when AI hallucinations compromise the accuracy of legal filings.
As AI tools become more prevalent in legal research and drafting, the judiciary may increasingly rely on existing procedural rules to address misconduct. The outcome of this case may serve as a precedent for future disputes involving AI‑generated legal content.
The sanctions imposed by Judge Aycock represent the first instance in the Fifth Circuit where a federal judge removed all attorneys from a case due to AI‑related misconduct. The decision may prompt other courts to adopt similar measures and could influence how law firms manage AI tools in the future.
