The Thomson Reuters logo is seen on the company building in Times Square, New York, USA, on January 30, 2018.REUTERS/Andrew Kelly/File Photo Obtaining license rights
Sept 25 (Reuters) – Information services firm Thomson Reuters (TRI.TO) illegally copied content from legal research platform Westlaw to train Ross Intelligence’s competing artificial intelligence-based platform A jury will have to decide the outcome of the lawsuit accusing them of doing so. A federal judge in Delaware said Monday.
of decision U.S. Circuit Judge Stefanos Vivas sets the stage for what could be one of the first trials related to the misuse of data to train AI systems. Tech companies such as Metaplatform (META.O), StabilityAI, and Microsoft-backed OpenAI (MSFT.O) have also been sued by writers, visual artists, and others over the use of copyrighted material to train their generative AI software. Facing lawsuits from other copyright owners. .
A spokeswoman for Thomson Reuters, the parent company of Reuters News, said it looked forward to presenting the evidence to the jury.
“This case continues to concern Mr. Ross’ theft of Thomson Reuters’ original commentary, analysis and organizational systems,” the spokesperson said. “We have sought summary judgment on selected issues because we believe the facts of the case are clear.”
Representatives for Mr. Ross did not respond to requests for comment.
In a 2020 Thomson Reuters lawsuit, legal research firm Ross Intelligence was accused of copying Westlaw’s “headnotes” summarizing the legal points of the court’s opinion. Thomson Reuters accused Ross of misusing thousands of headnotes to train an AI-based legal search engine.
Ross said Shut down the platform January 2021, citing costs of “sham” litigation. Reuters could not confirm whether it did so.
The companies asked the court to win the case pretrial. Mr. Ross argued in part that he was making fair use of Westlaw’s material and that it raises issues that could be pivotal in legal disputes over generative AI training.
Ross said the headnote material is used “as a means of finding judicial opinions” and that the company does not compete in the market for the material itself. Thomson Reuters responded that Ross copied the materials to build a direct competitor to Westlaw.
Vivas said Monday that other issues, including fair use and the scope of Thomson Reuters’ copyright protection in the headnotes, should be decided by a jury. He noted that the fair use analysis had win-win elements.
The judge said he could not determine whether Mr. Ross had “transformed” Westlaw’s materials into “an entirely new research platform that served a different purpose,” which is often a key question of fair use.
Mr. Vivas also said he could not decide whether a decision against Mr. Ross or Thomson Reuters would best serve the public interest.
“Here we come across a hotly debated issue,” Vivas said. “Is it in the public interest to allow AI to be trained using copyrighted material?”
No trial date has been set in this case.
Reporting by Blake Britten in Washington Editing by David Barrio and Lisa Shoemaker
Our standards: Thomson Reuters Trust Principles.
