Law Firms Face Malpractice Risks as AI Hallucinations Hit Courtrooms

Law firms are increasingly integrating generative AI into their research and drafting workflows, yet this push for efficiency is clashing with traditional mandates of candor and competence. According to data documented by The Washington Post, at least 95 U.S. court incidents involving AI-generated errors have occurred since June 2023. These failures, ranging from fabricated cases to misrepresented precedents, are now triggering severe judicial sanctions, including the removal of counsel and mandatory disclosures to state bars.

Crawford Appleby, a partner at Wisner Baum, argues that the crisis stems from firms viewing AI as a substitute for human diligence rather than a subordinate tool. A Stanford-led study highlighted the gravity of the situation, finding that leading legal AI research tools hallucinated between 17% and 33% of the time under test conditions. To mitigate these risks, Wisner Baum has implemented strict internal governance, including mandatory human verification through trusted databases like Westlaw or Lexis and the development of in-house tools designed specifically to identify machine-generated errors.

The American Bar Association’s Formal Opinion 512 underscores that the responsibility for court filings remains entirely with the human attorney, regardless of the software used. As courts grow less tolerant of technical excuses, firms are being urged to abandon the rush toward unbridled AI adoption in favor of a measured approach. For practitioners, accuracy is no longer just a matter of administrative housekeeping; it is a critical component of maintaining judicial integrity and avoiding malpractice exposure.