Scholars Agree Opinion in Thomson Reuters v. Ross Should Be Disregarded

The Disruptive Competition Project is taking part in Fair Use/Fair Dealing Week, an annual celebration of the important doctrines of fair use and fair dealing.

A Third Circuit appeals court judge sitting by designation in Delaware district court recently issued a controversial opinion in Thomson Reuters v. Ross, revising much of his previous decision to grant summary judgment, with holdings on copyright and fair use that conflict with established precedent. While this case addressed machine learning rather than generative AI explicitly, consensus from copyright experts including leading scholars has overwhelmingly detailed their many concerns with the ruling and its approach to the fair use doctrine and the copyrightability of headnotes and snippets. They agree that the opinion should not be followed in the numerous ongoing cases on generative AI currently progressing in various courts. 

While this was just one district court decision in one U.S. state, its complex procedural posture—and the fact that the defendant’s service has since shut down due to legal costs—leave the path to appeal uncertain. As policymakers and courts grapple with the nuanced intersection between copyright and AI, they should not rely on this flawed decision, which runs contrary to U.S. copyright law and the objective of American innovation and competitiveness in the artificial intelligence space.

Fortunately, U.S. courts’ interpretation of limits to the scope of copyright protection including the fair use doctrine has promoted innovative new technology and free expression, which we celebrate during Fair Use Week. Policymakers would be wise to listen to the many leading voices explaining why this particular decision is such an outlier:

Professor Eric Goldman laid out many elements of the opinion that he disagrees with:

“First, the court ignores cases that applied the transformativeness standard to content indexing–most importantly, the Google Books ruling–and instead bizarrely focused on cases involving software code…I don’t see how the court can say there was no ‘transformation’ when Ross’ outputs included none of the allegedly infringing material. It’s impossible to be more transformative than that.” 

Professor Matthew Sag questioned the ruling’s analysis of transformativeness: 

“The court’s conclusion that Ross’s use was not transformative is puzzling, especially given its acknowledgment—while discussing the third fair use factor—that the output of Ross’s system did not replicate Westlaw’s copyrighted headnotes but rather produced uncopyrighted judicial opinions.” 

“But what, exactly, sets this apart from generative AI? More broadly, how does this differ from other cases where nonexpressive uses have been deemed fair use? The opinion offers little guidance. It fails to engage with seemingly comparable precedents, such as plagiarism detection tools, library digitization for text analysis and digital humanities research, or the creation of a book search engine—cases where courts have found fair use.”

Authors Alliance detailed the consequences if the case were widely accepted: 

“The impact of this case is currently limited, both because it is a district court ruling and because it concerns non-generative AI. However, it is important to remain vigilant, as the reasoning put forth by the ROSS court could influence other judges, policymakers, and even the broader public, if left unchallenged.”

“This ruling combines several problematic arguments that, if accepted more widely, could have significant consequences. First, it blurs the line between fact and expression, suggesting that factual information can become copyrightable simply by being written down by someone in a minimally creative way. Second, it expands copyright enforcement to intermediate copies, meaning that even temporary, non-public use of copyrighted material could be subject to infringement claims. Third, it conjures up a new market for AI training data, regardless of whether such a licensing market is legitimate or even likely to exist.” 

Re:Create’s Brandon Butler emphasized the need to protect consumers: 

“Judge Stephanos Bibas compared this process to a sculptor who removes marble to create an original statue, but it’s closer to the work done by technicians who fine-tune 3D scans: a process whose goal is not to express something original but rather to faithfully reproduce something that already exists. That’s not the kind of work copyright protects, and that should have been a bigger factor in this case.”

“Content oligarchs like Thomson Reuters have enough copyrights to threaten ruination on anyone who tries to develop a technology that touches their market. Policymakers should be looking for ways to protect creators, competitors, and consumers from these shakedowns, rather than ways to make the grift even easier.” 

Professors Mark Bartholomew and Harry Surden noted the important nuances in copyright and AI cases in Business Insider:

Prof. Bartholomew “does not think the ruling will have a dramatic effect on the other major AI-related copyright cases since it’s just one opinion of a lower court and ‘neglects some of the most important fair use case law.’” 

Prof. Surden “fears judges overseeing other AI copyright lawsuits may not take into account the differences between the cases. The ruling is ‘likely to confuse judges and others on the issue of generative AI, but there are significant differences that are very subtle and most people will miss,’ Surden said, explaining, ‘It’s both a different technology and the company was acting kind of duplicitously here’…’Judges tend not to be experts in copyright law or generative AI, so I think they’re likely to miss a lot of these nuances, and I think it will, unfortunately, have an effect on other cases,’ said Surden.”

Professor James Grimmelmann highlighted potential harms to generative AI companies in Wired:

Prof. Grimmelmann: “‘If this decision is followed elsewhere, it’s really bad for the generative AI companies.’ Grimmelmann believes that Bibas’ judgment suggests that much of the case law that generative AI companies are citing to argue fair use is ‘irrelevant.’”

Overall, this ruling has sparked widespread criticism from copyright scholars and academics.  As AI-related copyright cases continue to be considered, courts and policymakers must maintain the balance that’s fundamental to U.S. law. For Fair Use Week, we celebrate this essential law that, when properly interpreted, ensures copyright remains a tool for advancement and creativity.