Three attorneys are in hot water with the State Bar of California after serious allegations regarding their use of artificial intelligence in court documentation. Accusations have emerged that these legal professionals cited non-existent cases in their submissions, raising significant concerns over ethical practices in the legal field.
Disciplinary Charges Filed
The State Bar of California has officially filed disciplinary notices against Omid Emile Khalifeh, an attorney in Los Angeles, and Steven Thomas Romeyn, who practices in Scottsdale, Arizona. Both are accused of misusing AI to draft court filings. It’s important to note that no ruling has yet been made by the State Bar Court on these allegations.
In another related case, Beverly Hills attorney Sepideh Ardestani faced sanctions for including erroneous and non-existent citations in a federal filing made in March 2025. None of the attorneys could be reached for immediate comment regarding these charges.
Guidelines on AI in Legal Practices
While California attorneys are permitted to utilize generative AI tools to create legal documents, they bear the responsibility of ensuring the information they present is accurate. Recently, a select group of judges in Los Angeles County have begun using an AI tool called Learned Hand to assist in summarizing motions and drafting rulings, highlighting the incorporation of technology in the judicial process.
Chief trial counsel George Cardona emphasized that these incidents reflect a broader issue of trust within the legal system when attorneys fail to verify the accuracy of their submissions. AI tools, while innovative, can generate erroneous information. “Courts and clients must be able to trust that the filings attorneys submit are accurate and compliant with professional standards,” he stated.
Khalifeh has been linked to six allegations of misconduct related to his use of AI in a trademark case, including the submission of citations that were fabricated, along with violations of a standing court order requiring disclosure of AI usage. He maintains that all the citations he included came from legitimate judicial decisions.
In a separate case, Romeyn faces scrutiny for submitting similarly flawed citations in a personal injury case. When his work was questioned, he admitted to using AI, stating he did review several citations but neglected to verify each one before submitting them.
The outcomes for Khalifeh and Romeyn hinge on the State Bar Court’s ruling regarding their professional conduct, which could lead to suspension or disbarment as determined by the California Supreme Court.
Meanwhile, Ardestani’s case concluded with the acknowledgment of her inadequate disclosure about the use of fictional citations in a wage-and-hour class-action lawsuit. Although she did not admit to utilizing AI, her cited errors reportedly stemmed from taking notes on other matters.
The Eastern District of California criticized the time spent addressing Ardestani’s misconduct, labeling it an unnecessary drain on judicial resources. Her disciplinary stipulation includes a year-long probation, a 30-day license suspension, and mandatory continuing legal education focusing on technology and the implications of AI in legal work.
