The Intercept’s Lawsuit Against OpenAI Advances on Claim It Removed Reporters’ Bylines

In an order on Friday, a federal court rejected OpenAI’s effort to toss a lawsuit filed by The Intercept over using its journalists’ work to train ChatGPT without permission or credit.

One of The Intercept’s claims under the Digital Millenium Copyright Act will move forward against OpenAI, although another claim was dismissed, along with all of its claims against Microsoft. 

“This decision shows that the DMCA provides critical safeguards for news organizations against encroachment by AI companies and presents a viable approach to challenging the unauthorized use of digital news content for AI training, regardless of whether that content has been registered with the Copyright Office,” said Matt Topic, The Intercept’s attorney in the litigation and a partner at law firm Loevy + Loevy, in a statement following the order.

This is a “first-of-its kind decision,” according to Topic.

The Intercept filed its lawsuit against OpenAI and Microsoft in February, joining a growing number of media outlets that accuse the companies of various copyright violations. The Intercept offered a novel argument under a provision of the DMCA that forbids stripping out “copyright management information” such as a work’s title and author, which The Intercept alleges OpenAI did in building the training data for ChatGPT.

In a brief order, Judge Jed S. Rakoff of the U.S. District Court for the Southern District of New York allowed this claim to proceed against OpenAI. Rakoff wrote that a full opinion “explaining the reasons for this ruling will issue in due course.”

Rakoff dismissed The Intercept’s claim that OpenAI violated another provision of the DMCA, which prohibits distributing content knowing that this copyright management information was removed. He also dismissed The Intercept’s claims against Microsoft under both DMCA provisions.