That's part of libel plaintiff's argument(filed last week) for subpoenaing defendant's AI queries in Fontainebleau Florida Hotel, LLC v. Botach; the defendant is popularly known as Rabbi Shmuley Boteach. From the introduction:
The Court asked whether Botach's communications with AI platforms fall within the scope of "damages" discovery permitted by ECF No. 261, and specifically how those materials relate to malice and the statutory cap on punitive damages. Precedent from the United States Supreme Court and Eleventh Circuit, as well as Florida law, answers that question directly.
Under Fla. Stat. § 768.73(1)(c), the cap on punitive damages lifts entirely where the fact finder determines that "the defendant had a specific intent to harm the claimant" at the time of the injury. As discussed below, legions of cases from trial courts to the Supreme Court hold that post-defamation conduct is probative of whether defamatory statements were made with intent to harm. Specific intent is therefore not background color to the damages trial or damages discovery. Discovery into Botach's intent to harm Plaintiff is damages discovery, and the AI materials are the most direct evidence of that intent.
Botach did not merely publish an errant falsehood on a blog or make a heated statement in a debate. He methodically engineered his attacks—drafting, refining, and iterating for painful effect through the most potent technologies known to humankind. The prompts and chats are his mental state reduced to writing, and they are probative of damages.
And from the defendant's response:
