A federal judge in Nevada has ruled that a state court judge accused of "relying wholly on artificial intelligence" to decide a case cannot be sued over it — because issuing a ruling, even an allegedly AI-generated one, is a judicial act protected by absolute immunity. The decision, spotlighted Monday by legal scholar Eugene Volokh on Reason's Volokh Conspiracy blog, adds an unexpected chapter to the growing collision between AI and the courts. Follow our AI news coverage for the latest on AI policy and law.
The Case: Phillips v. Parlade For more context on this story, see our ongoing AI trends.
The ruling came from Judge Gloria Navarro of the U.S. District Court for the District of Nevada in Phillips v. Parlade, issued last Wednesday. A litigant sued the state court judge who had presided over his case, arguing that she had effectively outsourced her job to a machine.
"Plaintiff … argues that judicial immunity does not apply in this matter because Defendant unlawfully delegated her official decision-making duties when she relied wholly on artificial intelligence to issue a judicial ruling, without any discretionary human thought, such that her actions cannot be considered a 'judicial act,'" the decision recounts. The plaintiff further argued that delegating "100%" of her decision-making meant the rulings were issued in a clear absence of all jurisdiction.
The court was unmoved. Citing long-standing precedent, Judge Navarro reaffirmed that "judges enjoy absolute immunity from civil liability, even if their action was in error, done maliciously, or in excess of their authority." The immunity yields only in two narrow situations: when the challenged conduct is accompanied by a clear absence of all jurisdiction, or when it is not judicial in nature.
Why the Claim Failed
To determine whether an act is judicial, courts apply a four-factor test: whether the act is a normal judicial function, whether the events occurred in the judge's chambers, whether the controversy centered on the case pending before the judge, and whether the events arose out of a confrontation with the judge in her official capacity.
"Here, Plaintiff alleges that Defendant issued a judicial decision in his state court case by relying wholly on artificial intelligence," the decision states. "Issuing a judicial ruling is clearly a normal judicial function and the controversy at issue centered on Plaintiff's state court case pending before Defendant." Finding no allegations placing the events outside chambers and no legal authority supporting a finding of a clear absence of jurisdiction, the court held the defendant entitled to judicial immunity and dismissed the case.
Notably, the ruling does not find that the state judge actually used AI — only that even if the allegations were true, immunity would still shield her. As Volokh, a senior fellow at the Hoover Institution and emeritus UCLA law professor, put it: the federal decision means that "as a matter of law, even if the allegations are correct and she had indeed relied entirely on AI in making her decision, she can't be sued for that in federal court."
The Remedies That Remain
The decision does not leave the hypothetical AI-dependent judge unaccountable — it just moves the accountability elsewhere. Objections to a state judge's reliance on AI can be raised on direct appeal, through appeal-like remedies such as petitions for a writ of mandamus, or in state judicial disciplinary proceedings, Volokh noted. What they cannot become is a federal damages lawsuit against the judge personally.
The ruling landed as courts across the country continue to grapple with AI in litigation. In one of the latest examples, an appeals court this month flagged AI-generated fake case citations in a San Antonio school district lawsuit, KSAT reported — part of a steady stream of hallucinated-authority incidents that has prompted many courts to adopt AI disclosure orders for attorneys.
The Phillips v. Parlade decision points at a different, less examined problem: not lawyers citing fake cases, but judges allegedly delegating real ones. Absolute judicial immunity was designed to protect rulings from second-guessing by angry litigants. As Monday's discussion on Hacker News showed — where the story drew dozens of comments — many readers were unsettled that the doctrine extends, at least on paper, to a ruling a judge supposedly issued without any discretionary human thought at all.
It is worth stressing what the decision is not. It is not a finding that any judge has actually delegated a ruling to AI — the complaint's allegation remains unproven, and the federal court took it as true only for the purpose of the immunity analysis. And as a single district court ruling, it does not bind other courts. But it demonstrates how readily a doctrine built for human misconduct absorbs allegations of machine delegation: the four-factor test asks what the judge did, not how the judge thought. Under that framework, the identity of the author — human, AI, or something in between — simply never comes into view.
That blind spot may matter more as AI assistance spreads through courthouses. The plaintiff's theory was that a ruling produced "without any discretionary human thought" ceases to be a judicial act at all. The Nevada court's answer — that issuance of a ruling is "clearly a normal judicial function" regardless — leaves future litigants to argue about where, exactly, human judgment ends. The decision gives judges broad comfort that experimenting with AI tools will not expose them to personal liability, while giving litigants one fewer avenue to challenge outcomes they believe a machine produced.
For now, the answer in Nevada is clear: the place to challenge an AI-written ruling is the appellate docket, not the courthouse door for damages.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
