OpenAI Achieved a Legal Victory Against Progressive Publishers—Yet the Battle Continues
A case against OpenAI initiated by Alternet and Raw Story was dismissed by a judge, which may represent an important decision in the ongoing conflict between AI firms and publishers.
OPENAI HAS ACHIEVED a win in its ongoing legal battle with publishers regarding the use of creative content by its AI tools. On November 7, a judge threw out a copyright lawsuit against the startup filed by independent publishers Alternet and Raw Story.
Although numerous publishers have arranged content agreements with OpenAI, many copyright lawsuits targeting AI startups are progressing through the US legal system. Numerous complaints claim direct copyright violations, asserting that AI firms unlawfully train their systems using news articles, books, paintings, and other copyrighted content without authorization. Some also incorporate additional allegations that range from breaches of trademark law to infringements of the Digital Millennium Copyright Act, a copyright statute designed as an anti-piracy measure that is now widely used by intellectual property rights owners.
The grievance filed by Alternet and Raw Story centered around the DMCA, claiming that OpenAI violated the law by harvesting thousands of news articles and removing “copyright management information” (CMI) such as the author’s name, usage terms, and the title of the piece. The plaintiffs sought statutory damages of at least $2,500 for each violation, contending that OpenAI was aware that removing CMI from the training data would lead to copyright violations when ChatGPT summarized or “reproduced” articles without adequate attribution.
OpenAI contended that the publishers lacked legal grounds to file this claim, asserting they did not provide evidence that ChatGPT was trained on their content, much less that the training caused harm. Judge Colleen McMahon of the Southern District of New York concurred with OpenAI’s position, dismissing the lawsuit due to insufficient standing.
“According to OpenAI representative Jason Deutrom, ‘We develop our AI models utilizing data that is publicly accessible, in a way that adheres to fair use and similar principles, while being backed by established and broadly recognized legal precedents.'”
While this represents a significant obstacle for Alternet and Raw Story, it doesn’t have to be the conclusion. “We plan to pursue the case,” states John Byrne, founder and CEO of Raw Story. The following step involves seeking the judge’s approval to submit a revised complaint.
“Matt Topic, a partner at Loevy & Loevy, the firm representing Raw Story Media, states, ‘We believe we can resolve the court’s concerns in a revised complaint.'” Although Judge McMahon characterizes herself as “skeptical” that the outlets could “claim a recognizable harm” in the dismissal, her decision suggests that she is willing to entertain a new submission.
Topic, who similarly advocates for The Intercept in a related DMCA case involving OpenAI, along with the nonprofit newsroom the Center for Investigative Reporting in a copyright infringement lawsuit against OpenAI and Microsoft, expresses that he is “confident these types of DMCA claims are allowed by the Constitution.”
Not all specialists concur. “Those assertions are illogical and ought to be rejected entirely, so I’m not taken aback by this decision,” states Matthew Sag, a legal and artificial intelligence professor at Emory University. He thinks the publishers were unable to demonstrate that OpenAI violated the law partly because they did not provide specific instances showing that ChatGPT shared copies of their work after removing CMI.
Ann G. Fort, a partner and intellectual property attorney at Eversheds Sutherland, believes that the media organizations will have to present concrete instances of how ChatGPT generates infringing replies. “They will have to demonstrate results,” she states.
DMCA claims have proven to be particularly controversial in various AI legal cases. In The Intercept case, OpenAI submitted a motion to dismiss based on standing as well, but the court’s process was somewhat different, and the publisher was permitted to submit a revised complaint. It accomplished this last summer, strengthening its position by including 600 pages of exhibits, which featured instances of how OpenAI’s models could be instructed to generate text fragments that resembled an Intercept article in at least one case. The court is anticipated to make a decision later this month.
Regardless of whether Raw Story and Alternet are ultimately permitted to submit an amended complaint, this week’s dismissal does not seem to eliminate other legal arguments; the judge explicitly remarked that she found the particular DMCA claims deficient rather than the more general idea of infringement. “Let us clarify what is truly at risk here.” Judge McMahon writes, “What the plaintiffs genuinely seek compensation for is not CMI’s removal from the defendant’s training sets, but the defendant’s utilization of the plaintiff’s articles to create ChatGPT without compensating the plaintiff.” “It’s yet to be determined if another statute or legal theory would raise this kind of harm.” “However, that issue is not being addressed by the court today.” Nonetheless, certain experts think this decision might, in fact, result in significant repercussions. “This theory regarding the absence of standing could be a significant upheaval that extends well beyond AI,” states James Grimmelmann, a professor specializing in digital and internet law at Cornell University. “It has the ability to notably limit the types of IP cases that federal courts are able to consider.” He believes the reasoning used in this situation might be broadened to claim that publishers lack the standing “to litigate over model training entirely, even regarding copyright violation.”
