Three California attorneys are under scrutiny after state bar authorities accused them of incorporating fabricated and inaccurate legal citations into court documents using generative AI, igniting a broader debate on AI regulation within the legal profession.
Three California lawyers are facing discipline after state bar authorities said artificial intelligence helped produce court filings that contained invented or inaccurate legal citations, deepening a debate over how the legal profession should police the use of generative tools. The Los Angeles Times reported that the State Bar of California has brought charges against Los Angeles attorney Omid Emile Khalifeh and Scottsdale-based Steven Thomas Romeyn, while a separate disciplinary agreement has already been approved for Beverly Hills lawyer Sepideh Ardestani. California allows lawyers to use generative AI in drafting, but not to outsource the duty of checking every citation and factual claim before a filing goes to court.
According to the State Bar, Khalifeh faces six misconduct charges over an April filing in a federal trademark case in Los Angeles that allegedly included one nonexistent decision and two others that were irrelevant to the arguments being made. He initially told the court he had reviewed and verified the brief after using Lexis+ AI, but when judges pressed the issue he conceded he could not confirm one citation existed and removed it. Romeyn is accused of submitting irrelevant and fabricated citations in an October filing in an Orange County personal injury case, later acknowledging that he had not checked every citation before filing. The State Bar Court has yet to decide whether either lawyer committed professional misconduct.
Ardestani’s case has already moved further. The State Bar Court approved a stipulation on 6 April calling for one year of probation, including a 30-day suspension and mandatory technology-focused continuing legal education, with at least five hours devoted to the benefits and risks of AI in legal work. She did not admit using AI, but the court said she failed to back up her explanation that the false citations came from handwritten notes taken in another matter. The Eastern District of California described the episode as a drain on limited judicial resources, a comment that reflects growing irritation in courts confronting AI-generated errors.
Chief trial counsel George Cardona said the cases show what happens when lawyers fail to verify material generated with AI, noting that courts and clients must be able to rely on filings as accurate and professionally sound. California Courts reporting has said judges across the state are increasingly wrestling with sanctions, fines and probation recommendations tied to fabricated citations and other AI “hallucinations”, even as the broader national approach to discipline remains unsettled. The State Bar says the Supreme Court of California will decide whether any recommended punishment should be imposed, including possible suspension or disbarment.
Source Reference Map
Inspired by headline at: [1]
Sources by paragraph:
Source: Noah Wire Services
Noah Fact Check Pro
The draft above was created using the information available at the time the story first
emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed
below. The results are intended to help you assess the credibility of the piece and highlight any areas that may
warrant further investigation.
Freshness check
Score:
8
Notes:
The article was published on April 13, 2026, and reports on recent disciplinary actions taken by the State Bar of California against three attorneys for submitting court filings with fabricated citations generated by AI. ([latimes.com](https://www.latimes.com/california/story/2026-04-13/attorneys-used-ai-cited-fake-legal-decisions-state-bar-alleges?utm_source=openai)) Similar reports have appeared in other reputable sources, such as Hoodline and News Minimalist, indicating that the information is current and not recycled. ([hoodline.com](https://hoodline.com/2026/04/ai-hallucinations-put-three-california-lawyers-in-state-bar-crosshairs/?utm_source=openai))
Quotes check
Score:
7
Notes:
The article includes direct quotes from Chief Trial Counsel George Cardona, stating, “Courts and clients must be able to trust that the filings attorneys submit are accurate, supported, and compliant with professional standards.” ([latimes.com](https://www.latimes.com/california/story/2026-04-13/attorneys-used-ai-cited-fake-legal-decisions-state-bar-alleges?utm_source=openai)) A search for this quote reveals it is used in multiple sources, suggesting it may be a standard statement from the State Bar. The exact origin of the quote is not clear, raising questions about its originality.
Source reliability
Score:
9
Notes:
The article is published by the Los Angeles Times, a major and reputable news organisation. The information is corroborated by other reputable sources, such as Hoodline and News Minimalist, indicating a high level of reliability. ([hoodline.com](https://hoodline.com/2026/04/ai-hallucinations-put-three-california-lawyers-in-state-bar-crosshairs/?utm_source=openai))
Plausibility check
Score:
8
Notes:
The claims about attorneys facing discipline for submitting AI-generated court filings with fabricated citations are plausible and align with recent discussions on the use of AI in legal practice. Similar cases have been reported, such as the State Bar of California admitting to using AI to develop exam questions, which sparked controversy. ([latimes.com](https://www.latimes.com/california/story/2025-04-23/state-bar-of-california-used-ai-for-exam-questions?utm_source=openai))
Overall assessment
Verdict (FAIL, OPEN, PASS): PASS
Confidence (LOW, MEDIUM, HIGH): MEDIUM
Summary:
The article provides current and plausible information about disciplinary actions taken by the State Bar of California against attorneys for submitting AI-generated court filings with fabricated citations. While the source is reputable, the exact origin of some quotes is unclear, and the independence of verification sources is somewhat compromised due to aggregation. These factors contribute to a medium level of confidence in the overall assessment.
