Last week, a The federal judge in New York ruled a key copyright infringement claim by The Intercept against OpenAI would move forward in court. The ruling is the latest in a series of major legal decisions involving the AI ​​developer this month, after OpenAI tried to fend off lawsuits from several digital news publishers.

Judge Jed Rakoff said he would hear the claim that OpenAI removed author information when he allegedly fed The Intercept’s articles into the training datasets he used to build ChatGPT. Doing so could be a violation of Digital Millennium Copyright Act (DMCA)a 1998 law that, among other protections, makes it illegal to remove the author’s name, terms of use, or title from a digital work.

The judge rejected The Intercept’s claim that OpenAI had children knowingly distributed its articles after removing the DMCA-protected information. The judge also dismissed all of The Intercept’s claims against Microsoft, which has a billions of dollars investment in OpenAI and was named in the original filing. An opinion of the judge will be published in the coming weeks, setting out his reasoning for the dismissals.

“The decision allows a DMCA claim by digital publishers who do not have copyright records to proceed against OpenAI,” said Matte subjectpartner at Loevy & Loevy, which represents The Intercept. “We’re obviously disappointed to lose the claims against Microsoft, but the underlying complaint is the DMCA complaint against OpenAI, and we’re very happy to see that continue.”

“Our models are trained on publicly available data based on fair use and related principles that we believe are fair to creators,” OpenAI spokesman Jason Deutrom said in a statement.

At the beginning of this year I reported that The Intercept case created a new legal strategy for digital news publishers to sue OpenAI.

The New York Times lawsuit against OpenAI and similar lawsuits filed by New York Daily News and Mother Jonesleads with claims of copyright infringement. Lawsuits for infringement require that the relevant work first be registered with the US Copyright Office (USCO). But most digital news publishers don’t have archives of registered articles. For many, including The Intercept, submitting all of their Internet-published work to the USCO is too expensive or burdensome.

Until this summer, the government body required that each individual website article page be submitted and charged separately. In August, however, the USCO added a rule which allows “news websites” to submit articles in bulk. Among other reasons, the decision cited concerns about the unchecked infringement of online news content and the hope that copyright registrations would remain “adaptable to technological change.” But for most digital news publishers seeking legal action against OpenAI, particularly for using their work to train ChatGPT, the new rule came too late.

So far, The Intercept’s case is the only non-copyright litigation by a news publisher to make it past the motion-to-rescind stage.

Earlier this month, legal strategy focused on the DMCA received a major blow when another federal judge in New York dismissed all DMCA claims against OpenAI filed by Raw Story and AlterNet. Progressive digital news sites are jointly represented by Loevy & Loevy.

“Let’s be clear about what is really at stake here. The alleged harm for which plaintiffs truly seek damages is not the exclusion (of content management information) from defendants’ training kits, but rather defendants’ use of plaintiffs’ articles to develop ChatGPT without compensation,” Judge Colleen MacMahon wrote. in that decision.

Despite the setback, the judge said he would consider an amended complaint against OpenAI that would take her concerns into account. A proposed complaint by Raw Story and AlterNet was filed by Loevy & Loevy last week, even before The Intercept’s ruling was announced.

“When they supplemented their training sets with journalistic works, the defendants had a choice: they could train ChatGPT using journalistic works with the DMCA-protected copyright management information intact, or they could remove them. The defendants chose the latter,” it says proposed amended complaint. “In the process, (OpenAI) instructed ChatGPT not to recognize or respect copyright, not to notify ChatGPT users when responses they received were protected by journalists’ copyrights, and not to provide attribution when I use the work of human journalists.”

Like The Intercept, Raw Story and AlterNet are seeking $2,500 in damages for each instance in which OpenAI allegedly removed DMCA-protected information from its training datasets. If damages are calculated based on each article allegedly used to train ChatGPT, it could quickly reach tens of thousands of violations.

“The proposed amended complaint would match and probably even exceed the surviving allegations in The Intercept case,” Topic said. “Different judges may come up differently on the same issue, but we are optimistic that we will be able to proceed with an amended motion.”

It’s unclear whether the Intercept ruling will encourage other publications to consider DMCA litigation; few publications have so far followed in their footsteps. As time goes on, there is concern that new lawsuits against OpenAI would be vulnerable to statute of limitations restrictions, particularly if news publishers want to cite the training datasets that underlie ChatGPT. But the decision is a signal that Loevy & Loevy is limiting itself to a specific DMCA claim that can actually stand up in court.

“We think the surviving claim for The Intercept is a claim that most digital publishers could make,” Topic said.

Unsplash