Law Firms Face Ethics Tests as AI Filing Errors Trigger Sanctions
AI errors in legal filings are resulting in high-profile sanctions and apologetic law firm admissions.
Why it matters: Legal practitioners integrating AI must adapt processes to avoid ethical pitfalls and professional discipline. Understanding these challenges is key for maintaining standards amid rapid tech adoption.
- Sullivan & Cromwell and Gordon Rees Scully Mansukhani LLP have both apologized for AI-generated citation errors in court filings.
- The U.S. Court of Appeals for the Sixth Circuit issued $30,000 in sanctions in March 2026 for AI-related mistakes.
- Several firms and lawyers have received fines ranging from $2,000 to $5,000 for submitting filings with AI-generated fake cases.
- Ethics experts highlight growing tension between traditional legal standards and the realities of powerful but fallible AI output.
The rapid adoption of AI in law firms is generating new ethical challenges, as high-profile incidents show the risk of over-reliance on imperfect technology. In recent court cases, lawyers have faced sanctions after submitting filings riddled with fabrications—products of AI tools known to generate plausible but fictitious legal content, or 'hallucinations.'
- In April 2026, Sullivan & Cromwell admitted to submitting a bankruptcy court filing with fabricated citations due to AI errors, mirroring a similar instance by Gordon Rees Scully Mansukhani LLP in October 2025.
- The U.S. Court of Appeals for the Sixth Circuit sanctioned two attorneys $30,000 in March 2026, citing improper reliance on AI for inaccurate legal citations.
- Lower fines, including $5,000 in Manhattan and $2,000 in Texas, underscore the breadth of the issue, with lawyers penalized for failing to verify AI-generated citations as required by legal ethics rules.
While the American Bar Association requires attorneys to ensure competence and diligence—even when using advanced tools—experts warn the technology’s allure can sideline prudence. “The rules haven't really changed, but we're just having trouble following them when the output of some AI is so seductively good,” said Lucian Pera, Legal Ethics Expert at Adams & Reese.
With reputational risks mounting and both state and federal judges increasingly skeptical, firms are under pressure to tighten oversight and verification processes for AI-assisted legal work.
By the numbers:
- $30,000 — Sanctions imposed by the U.S. Court of Appeals for AI citation errors (Mar 2026)
- $5,000 — Fine for citing fake AI-generated cases in Manhattan federal court (Jun 2023)
- $2,000 — Fine for submitting AI-generated fake cases in Texas (Nov 2024)