Judge Withdraws Opinion After Fabricated Quotes and Legal Errors Are Exposed

Bloomberg’s Justin Henry reports that a federal judge in New Jersey withdrew a ruling in a securities class action after attorneys revealed that it contained multiple factual and legal errors, including fabricated quotes and case misstatements.

The retracted opinion, issued by Judge Julien Xavier Neals in In Re CorMedix Inc. Securities Litigation, initially denied CorMedix Inc.’s motion to dismiss shareholder claims.

It was vacated on July 31, 2025, pending a corrected decision. The flawed opinion has sparked debate over judicial accountability and the role of AI in legal drafting.

According to Bloomberg, the case is a rare example of a judge being called out for the kind of legal drafting errors for which courts take lawyers to task. 

No direct mention of AI-assisted drafting was made, but the situation mirrors recent controversies over fake AI-generated citations in briefs, which have led to sanctions against lawyers.

The opinion drew scrutiny after attorneys from multiple cases flagged significant inaccuracies. These included invented quotations from case law, erroneous case citations, and misattributed statements.

For instance, Judge Neals attributed a quote to Dang v. Amarin Corp. that does not exist, and cited a Southern District of New York case that could not be found.

Although the attorneys did not formally seek reconsideration, the magnitude of the legal errors prompted the court to retract the decision. Neals has declined to comment.

The ruling’s impact extended beyond CorMedix. Plaintiffs in a related case against OutlookTherapeutics cited Neals’ original decision as supplemental authority to oppose a motion to dismiss.

In response, defense attorneys said the decision contained “pervasive and material inaccuracies,” and the Outlook plaintiffs withdrew their reliance on the CorMedix opinion.

The fact that a judge’s ruling has been undermined by errors that appear to be AI-related raises new questions about professional standards and judicial ethics. AI platforms routinely caution that they can and do make mistakes and need to be thoroughly checked.