close
close

OpenAI has won a legal victory over progressive publishers – but the fight is not over yet

Topic, who also represents The Intercept in a similar DMCA lawsuit against OpenAI and the nonprofit newsroom Center for Investigative Reporting in a copyright infringement lawsuit against OpenAI and Microsoft, says he is “confident that these types of DMCA claims will be allowed.” of the Constitution.”

Not all experts agree. “These claims make no sense and should all be dismissed, so this ruling doesn’t surprise me,” said Matthew Sag, a professor of law and artificial intelligence at Emory University. He believes the publishers failed to prove that OpenAI violated the law, in part because they did not provide specific examples of ChatGPT distributing copies of their work after CMI was removed.

Ann G. Fort, an intellectual property lawyer and partner at Eversheds Sutherland, suggests that news outlets need to provide concrete examples of how ChatGPT produces infringing responses. “You have to show results,” she says.

DMCA claims have been particularly contentious in a number of AI lawsuits. In The Intercept case, OpenAI also filed a motion to dismiss, but the court process was slightly different and the publisher received permission to file an amended complaint. This happened last summer, bolstering his argument by adding 600 pages of exhibits, including examples of how OpenAI's models could be made to produce snippets of text that, in at least one case, were nearly identical to an Intercept article. The court is expected to make a decision later this month.

Regardless of whether Raw Story and Alternet are ultimately allowed to file an amended complaint or not, this week's dismissal does not appear to preclude other legal arguments; The judge specifically noted that in her opinion the specific DMCA claims were missing, rather than the broader concept of infringement. “Let’s be clear about what this is really about. “The alleged harm for which plaintiffs are actually seeking redress is not the exclusion of CMI from defendant’s training sets, but rather defendant’s use of plaintiff’s articles to develop ChatGPT without compensation to plaintiff,” writes Judge McMahon. “Whether there is another law or legal theory that increases this type of harm remains to be seen. But that question is not before the court today.”

However, some experts believe that this ruling could actually have far-reaching consequences. “This theory of ineligibility is actually a potential earthquake that extends far beyond AI,” says James Grimmelmann, a professor of digital and internet law at Cornell University. “It has the potential to significantly limit the types of IP cases that federal courts can hear.” He suspects that the logic used in this case could be extended to mean that publishers do not have standing to “sue over model training, itself for copyright infringement.”