Nvidia Rebuffs Authors’ Claims It Trained AI on Pirated Books

Nvidia Corp. asked a court to toss a group of authors’ latest claims that the tech company accessed pirated content to train its artificial intelligence models.

The AI giant said the authors’ renewed legal complaint over alleged copyright infringement is “replete with vague references” to unidentified large language models and doesn’t show which datasets were used to train them, “let alone that the datasets contained their works,” according to a motion to dismiss the claims filed Thursday in the US District Court for the Northern District of California.

Authors including Abdi Nazemian expanded on their earlier allegations that Nvidia acquired ...