Skip to content

Individual Action

Walters v. OpenAI

Judgment EnteredNot a class action

Mark Walters v. OpenAI, L.L.C.

Where it stands

The first U.S. defamation case tried to judgment over a generative-AI 'hallucination.' ChatGPT falsely told a journalist that Walters had been sued for embezzling from a gun-rights organization. On May 19, 2025, the court granted OpenAI summary judgment on three independent grounds: a reasonable reader in the journalist's position could not have understood the output as stating actual facts about Walters (given OpenAI's accuracy disclaimers and the journalist's own knowledge that the output was wrong); Walters could show neither negligence nor actual malice; and he proved no recoverable damages. The court declined to apply strict liability to AI output. No Section 230 holding was made.

Key ruling

May 19, 2025 summary judgment for OpenAI: AI-generated output accompanied by accuracy disclaimers, in context, was not reasonably understood as a statement of fact; strict liability rejected in favor of a fault-based standard for AI defamation.

Note on this record

A parallel federal action (N.D. Ga., 1:23-cv-03122) was filed and later resolved procedurally. No public record confirms whether Walters appealed; treat appellate status as unknown rather than final.

The record

Court
Superior Court of Gwinnett County, Georgia
Case number
23-A-04860-2
Filed
2023-06-05
Defendants
OpenAI, L.L.C.
Plaintiffs
Individual radio host suing over an allegedly defamatory ChatGPT output
Technology at issue
ChatGPT (OpenAI)
Relief sought
General and punitive damages for libel
Counsel of record
John Monroe (plaintiff) · Gibson, Dunn & Crutcher LLP and Fellows LaBriola LLP (for OpenAI)

Claims pleaded

Sources (3)

This page describes a publicly filed case. This site is not counsel on this matter, nothing here is legal advice, and the allegations described are allegations unless a court has said otherwise.

Back to the docketClaim encyclopedia