The contentious legal battle between Thomson Reuters and Ross Intelligence has taken another procedural turn as U.S. Circuit Judge Stephanos Bibas issued a memorandum clarifying the rationale behind his recent decision to delay the trial. Initially set for May 12, the trial’s postponement allows Ross to pursue an interlocutory appeal concerning significant copyright and fair use considerations.
Judge Bibas, presiding in the U.S. District Court in Delaware, acknowledged the complexities of the issues at hand. He stated that although he stands by his February 2025 summary judgment opinion, which favored Thomson Reuters, the questions involved are inherently challenging. This admission led to his decision to authorize an interlocutory appeal, believing it would resolve the case more efficiently and potentially avoid a full trial.
The memorandum opinion outlines two crucial questions now heading to the 3rd U.S. Circuit Court of Appeals:
- The originality of Westlaw’s headnotes and Key Number system and their eligibility for copyright protection.
- The applicability of fair use in Ross’s utilization of Westlaw’s headnotes to develop a competing legal research tool.
The dispute underscores a broader conversation about the intersection of AI and copyright law, given that Ross Intelligence allegedly used copyrighted materials from Thomson Reuters’ Westlaw to train its AI legal research platform. Judge Bibas originally sided with Ross on fair use grounds in 2024, only to reverse his position in February 2025.
Judge Bibas highlighted that these issues involve “novel and difficult questions of first impression” for the 3rd Circuit. His decision to grant the interlocutory appeal reflects not just the substantive legal questions but also practical considerations regarding Ross’s financial constraints, with the costs of a full trial potentially debilitating for the now-defunct AI startup.
The appellate court’s decision could have lasting implications for the legal publishing industry and how AI technologies leverage copyrighted materials. As the case progresses, it stands to set influential precedents on the permissibility of using copyrighted legal data to train AI systems and delineate the boundaries of copyright protection for legal publishers.
The timeline for a resolution remains open-ended, but Judge Bibas’s memorandum acts as a prelude to what may be a defining moment in AI-related copyright litigation.
Further updates on this case can be found on LawNext.