COLUMBUS, Ohio — Two Ohio attorneys are facing disciplinary action after legal documents they submitted contained false information generated by artificial intelligence, highlighting growing concerns about the risks of using AI in the legal profession.
The Ohio Office of Disciplinary Counsel recently filed a complaint against Cleveland-area attorney William Norman, alleging that inaccurate statements created by an AI program were included in a legal filing. According to the complaint, Norman acknowledged that a non-lawyer member of his staff used an artificial intelligence tool that fabricated comments supposedly made by a prosecutor—remarks that were never actually spoken.
The case follows another recent disciplinary matter involving a different Cleveland-area attorney. Earlier this summer, the Ohio Board of Professional Conduct recommended an indefinite suspension after that attorney admitted using AI software that cited court cases that did not exist.
Legal ethics experts say AI-generated errors, often referred to as “hallucinations,” have become an increasingly common problem despite years of warnings to attorneys.
Heather Zirke, director of the Miller Becker Center for Professional Responsibility at the University of Akron School of Law, said she regularly sees lawyers across the country facing sanctions for filing court documents containing fabricated citations or inaccurate legal information generated by AI.
Ohio Disciplinary Counsel Joe Caligiuri said the issue is becoming more frequent nationwide, affecting not only attorneys but occasionally judges as well. However, he noted the problem remains most common among individuals representing themselves in court without legal counsel.
Delaware County Probate Judge David Hejmanowski, who serves on the Ohio Supreme Court’s Artificial Intelligence Task Force, said many self-represented litigants now rely on free AI programs instead of traditional legal research. While those tools often produce convincing legal language, they can also generate entirely fictional court decisions or quotations that appear authentic.
Without legal training, users may not recognize those errors before submitting documents to a court.
The consequences can be significant. Courts have issued sanctions totaling thousands of dollars, dismissed cases because of inaccurate AI-generated filings, and imposed disciplinary action against attorneys who fail to verify the information produced by AI programs.
Hejmanowski said attorneys have had ample warning about the dangers of relying on AI without carefully reviewing its work, making such mistakes increasingly difficult to excuse.
Experts also raised concerns about privacy. Uploading confidential legal documents into publicly available AI platforms could expose sensitive client information if lawyers do not understand how that data is stored or used.
Another long-term concern involves the legal system itself. If AI-generated false citations make their way into published court decisions without being caught, they could eventually be referenced in future cases, potentially undermining the accuracy of legal research.
Despite those risks, legal professionals do not believe artificial intelligence should be abandoned altogether. Instead, they say attorneys should use AI as a research or drafting aid while independently verifying every citation, quotation and legal conclusion before filing documents with the court.
Experts also encourage attorneys to be transparent with clients about how AI may be used in their cases and what safeguards are in place to protect confidential information.
