A federal judge has ruled against Ross Intelligence Inc. in a copyright dispute with Thomson Reuters, concluding that Ross’s use of Westlaw headnotes to train its AI legal research tool was not protected under “fair use.” The decision, reported by Bloomberg Law, is a key moment in the growing tension between AI innovation and intellectual property rights.
U.S. District Judge Stephanos Bibas, presiding in Delaware, sided with Thomson Reuters on two fair use factors: Ross’s purpose for using the headnotes and the harm caused to Westlaw’s market. “Ross took the headnotes to make it easier to develop a competing legal research tool,” Bibas wrote. “So Ross’s use is not transformative.”
The judge partially granted summary judgment for Thomson Reuters, finding Ross infringed on more than 2,000 headnotes. Questions about whether copyright protection has expired on some headnotes and potential infringement of another 5,000 will go to trial.
A Precedent for AI and Copyright
Filed in 2020, this lawsuit is one of the first to address AI-related copyright issues. Though Ross’s tool uses non-generative AI, the decision could influence future cases involving generative AI models, such as those by OpenAI and Meta.
Shortly after the ruling, music publishers suing AI developer Anthropic over copyrighted song lyrics filed notice of the opinion in a separate California case, underscoring its potential ripple effects.
Thomson Reuters, in a statement, celebrated the ruling, saying, “We are pleased that the court granted summary judgment in our favor and concluded that Westlaw’s editorial content created and maintained by our attorney editors, is protected by copyright and cannot be used without our consent,” Ross’s legal team has yet to comment.
The case, delayed from a 2024 trial date, raises significant questions for AI developers, as courts increasingly scrutinize the use of copyrighted material for training models. As Bloomberg Law noted, the decision marks a pivotal moment in shaping the legal boundaries of AI innovation.
