Judge advances digital publisher Ziff Davis’ ChatGPT copyright infringement claims

MANHATTAN (CN) — A federal judge on Monday advanced a trio of copyright claims brought by digital publisher Ziff Davis against ChatGPT maker OpenAI, while the artificial intelligence firm won dismissal of several others.

Ziff Davis, a global digital media company whose portfolio includes leading brands in technology, shopping, gaming and entertainment like Mashable, PCMag and IGN, accuses OpenAI of scraping its online content without authorization to train the artificial intelligence chatbot’s language technology.

Ruling on OpenAI’s motion to dismiss Ziff Davis’ first amended complaint, U.S. District Judge Sidney Stein refused to throw out three of the publisher’s copyright infringement claims in which the Mashable publisher accused OpenAI of distributing copies of its copyrighted works with the copyright management information removed, in violation of the Digital Millennium Copyright Act.

“In sum, Ziff Davis has stated a claim for contributory infringement because ‘knowledge of specific infringements is not required to support a finding of contributory infringement’ and, by alleging ‘widely publicized instances of copyright infringement’ and ‘numerous examples of infringing outputs in’ the first amended complaint, Ziff Davis plausibly alleged end-user infringement and that OpenAI possessed actual or constructive knowledge of this third-party infringement,” the judge wrote in his 15-page opinion and order.

Stein, a Bill Clinton appointee, did grant dismissal with respect to claim 4, for common law unjust enrichment, and on claim 5, for circumvention of technological measures, with regard to “robots.txt” files in its websites’ code, which instructed OpenAI’s GPTBot not to scrape Ziff Davis’ websites.

“Robots.txt files instructing web crawlers to refrain from scraping certain content do not ’effectively control’ access to that content any more than a sign requesting that visitors ‘keep off the grass’ effectively controls access to a lawn,” Stein wrote.

“At most, Ziff Davis alleges that OpenAI disregarded the instructions that were contained in robots.txt files,” the judge concluded. “This is not ’circumvention’ under the DMCA, and Ziff Davis’ claim fails for this reason as well.”

Ziff Davis argued its complaint that, despite explicitly blocking OpenAI’s GPTBot scraping software from its websites through “robots.txt,” lines of instruction code published by OpenAI to the publisher still saw a spike in the bot’s activity on some of its sites.

Stein granted in part and denied in part with respect to dismissal of Ziff Davis’ claims for trademark dilution, ruling that only Ziff Davis’ Mashable is famous enough to plead trademark dilution over the outputs from ChatGPT’s large language models, while concluding that Lifehacker, CNET, ZDNET, PCMag, BabyCenter or IGN are not.

Stein also granted OpenAI’s motion to stay in part with respect to certain causes of action and as to newer ChatGPT models that are not already involved in the pending multidistrict litigation also underway before Stein: namely, ChatGPT’s o1, o1-mini, o1-pro, GPT-4.1, GPT-4.5, o3, o3-mini, o4-mini and GPT-5 models.

“The burden on OpenAI of potentially expansive discovery into these models — which are not otherwise at issue in this MDL — would involve additional deponents and documents and would be significant, and the likely delay attendant to such an expansion of the scope of discovery in this MDL at this time would be contrary to the interests of the court, non-parties, and the public in a speedy resolution of the core copyright issues in this MDL,” Stein wrote.

During oral arguments last month, Los Angeles-based intellectual property attorney Guy Ruttenberg argued a stay would effectively require Ziff Davis to litigate its claims twice, first on remand back to Delaware court and then again later in Manhattan federal court.

“That’s very, very disruptive to our case,” he said.

OpenAI’s attorney Joseph Gratz, meanwhile, argued that if the additional Ziff Davis claims and models are not stayed, it would slow down the discovery deadlines in place for the multidistrict litigation.

He said the eight newer ChatGPT models at issue in the Ziff Davis amended complaint would involve a greater scope of custodians to be deposed, in addition to new training data and source code for those additional software models.

“We are sprinting to reach the finish line with respect to the current discovery schedule,” Gratz said Thursday. “There is real hope that we can get there. … I think that hope vanishes if we add another eight, nine models to this thing.”

Representatives for neither company immediately responded to requests for comment Monday.

Stein is presiding over the multidistrict litigation, with numerous copyright infringement lawsuits from authors, news organizations and publishers against OpenAI and Microsoft consolidated together in the Southern District of New York, formally titled In re: OpenAI, Inc. Copyright Infringement Litigation.

Follow @jruss_jruss