Inside This Story:

✨ Artificial intelligence on trial — Judge rules that OpenAI's ChatGPT didn’t act with "actual malice" in spewing out false info about a gun rights radio host.

⚖️ Landmark test dismissed — Legal observers say this is one of the earliest court tests for AI-generated misinformation.

🚨 Error warnings saved OpenAI — The company’s extensive disclaimers and user guidance tipped the scales in its favor.

By Samuel Lopez – USA Herald

May 20 (USA Herald) — OpenAI has scored a courtroom victory in Georgia, where a judge dismissed a defamation lawsuit that accused ChatGPT of fabricating a bogus legal case against radio host and gun rights advocate Mark Walters.

The Superior Court of Gwinnett County ruled that Walters failed to demonstrate OpenAI acted with "actual malice" — the critical threshold in defamation law when the subject is a public figure. Judge Tracie Cason’s ruling emphasized that the company clearly warns users that its artificial intelligence chatbot may produce errors.

"OpenAI's industry-leading efforts to reduce errors of this kind and its extensive warnings to users that errors of this kind could occur negate any possibility that a jury could find OpenAI acted with actual malice here," Judge Cason wrote in her opinion.

Walters, a syndicated talk radio host known for his advocacy of Second Amendment rights, filed suit in 2023 after ChatGPT responded to a journalist's question by inventing a lawsuit in which Walters was falsely named as a defendant in a fictional gun-related legal matter.

The journalist, who worked for AmmoLand.com, later realized the information was entirely fictitious and never published the claims. Walters, however, said the error was deeply defamatory and damaging to his reputation.

This case marked one of the first legal attempts to hold an AI company accountable for misinformation generated by its model. Legal analysts viewed the lawsuit as a bellwether test for how courts will handle speech created by non-human agents.

OpenAI responded to the suit by noting that ChatGPT is an evolving tool designed to provide useful assistance, not verified legal fact. The company pointed out that users are constantly warned of its limitations, including a clear disclaimer that the system "may occasionally produce incorrect or misleading information."

That distinction proved critical to the court’s decision.

Under the U.S. Supreme Court’s landmark ruling in New York Times Co. v. Sullivan, public figures must prove that false statements were made with "actual malice" — that is, with knowledge of their falsity or with reckless disregard for the truth.

Judge Cason determined that Walters did not meet this burden. While acknowledging that the AI's response was false, the court found no evidence that OpenAI intended to defame or even knew the statements were incorrect.

OpenAI’s legal team, led by Theodore Boutrous Jr. of Gibson, Dunn & Crutcher and Matthew Macdonald of Wilson Sonsini Goodrich Rosati, argued that imposing liability for AI-generated hallucinations would create a chilling effect on innovation.

"This is a victory not only for OpenAI but for responsible AI deployment as a whole," said one attorney familiar with the case.

The case, Walters v. OpenAI LLC, raised thorny questions about accountability for generative AI. Could a machine be legally liable? And if not, what responsibility lies with the developers?

So far, the courts are leaning toward shielding companies that provide reasonable disclaimers and show good-faith efforts to mitigate harm. Legal experts say this decision sets an early precedent — although future courts may draw harder lines as AI becomes more sophisticated and widespread.

For now, OpenAI’s proactive user education seems to have shielded it from legal fallout.

Walters’ attorney, John Monroe, said Monday that they were reviewing the court’s decision for possible appeal.

What Comes Next

OpenAI welcomed the ruling, stating, "We appreciate the court’s careful decision and its findings about our efforts to responsibly educate users and improve our models."

Still, this isn’t the last time generative AI will appear in court.

"This decision may slow the rush to the courthouse, but it won't stop it. The real challenge will be when AI makes errors that lead to real-world damages. Those cases will test the boundaries of tech law and ethics in far more consequential ways." - Samuel Lopez.

Explore More
🔗 Follow us on X @RealUSAHerald