Raw Story Media, Inc. v. OpenAI, Inc.

District court grants motion to dismiss plaintiffs’ claims under Digital Millennium Copyright Act, on basis that plaintiffs lack Article III standing for their claims for damages and injunctive relief, and denies plaintiffs’ motion for leave to replead.

Plaintiffs Raw Story Media Inc. and AlterNet Media Inc. brought an action under the Digital Millennium Copyright Act (DMCA) against OpenAI Inc., OpenAI GP LLC, OpenAI LLC, Open AI Opco LLC, OpenAI Global LLC and OpenAI Holdings LLC. Plaintiffs are news organizations that have published over 400,000 breaking news features, investigative news articles and opinion columns online. Defendants are interrelated organizations responsible for the AI service ChatGPT. Plaintiffs allege that ChatGPT is trained on text known as “training sets” and that these training sets range from collections of links to a “scrape” of most of the internet. In this scrape of the internet used to train ChatGPT, plaintiffs claim that thousands of their copyrighted works of journalism were caught, stripped of their copyright information, put into ChatGPT’s training sets and used to train ChatGPT. Plaintiffs claim that because their copyright information was removed before being used to train ChatGPT, ChatGPT did not learn to communicate that information when responding to user inquiries using plaintiffs’ copyrighted works. Plaintiffs allege that defendants’ removal of the copyright management information (CMI) from plaintiffs’ works prior to training ChatGPT violated Section 1202(b)(i) of the DMCA, and that plaintiffs are entitled to damages. Plaintiffs further claim that because prior versions of ChatGPT generated significant amounts of plagiarized content, there is a substantial likelihood that the current version of ChatGPT will reproduce plaintiffs’ copyrighted works without providing the copyright information contained in those works, and on this basis, plaintiffs seek injunctive relief. 

Defendants filed a motion to dismiss plaintiffs’ complaint in its entirety, arguing that plaintiffs lack Article III standing to pursue their damages and injunctive relief claims and, in the alternative, fail to state a claim under the DMCA. The district court agreed that plaintiffs do not have Article III standing. In particular, with respect to plaintiffs’ damages claim, the court explained that plaintiffs have not alleged a concrete injury because even though plaintiffs allege that their copyrighted works, without CMI, were used to train ChatGPT and remain in ChatGPT’s repository of text, plaintiffs have not alleged any actual adverse effects stemming from this alleged DMCA violation. With respect to plaintiffs’ injunctive relief claim, the court similarly found that plaintiffs have not alleged any concrete injury because plaintiffs fail to allege facts showing that the risk of ChatGPT reproducing plaintiffs’ work, absent CMI, is substantial. The court continued that, given the amount of information that is contained in the repository used to train ChatGPT, the likelihood that ChatGPT would produce content plagiarized from one of plaintiffs’ articles is remote, and plaintiffs have not alleged facts demonstrating otherwise. Because the court found that plaintiffs lack Article III standing, the court did not reach the question of whether plaintiffs plausibly alleged a claim. The court granted defendants’ motion to dismiss in its entirety and denied plaintiffs’ motion for leave to replead without prejudice. 

Summary prepared by Safia Gray Hussain and Jennifer Kahn.