Photo credit: www.theverge.com
Court Rules in Favor of Thomson Reuters in AI Copyright Case Against Ross Intelligence
On Tuesday, a significant ruling emerged from the US District Court of Delaware, where Judge Stephanos Bibas issued a partial summary judgment favoring Thomson Reuters in its ongoing copyright infringement legal dispute with Ross Intelligence, a legal technology startup. Initiated in 2020, this case is one of the earliest to address the legal ramifications related to the use of AI tools and their training processes, particularly concerning the use of copyrighted materials without authorization.
This lawsuit aligns with a series of similar legal challenges currently facing major players in the AI sector, such as OpenAI and Microsoft. These cases often hinge on the contentious issue of whether these AI systems can invoke a fair use defense regarding their utilization of copyrighted content.
Thomson Reuters, through its spokesperson Jeff McCoy, expressed satisfaction with the court’s decision. In a statement given to The Verge, McCoy noted:
“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. The copying of our content was not ‘fair use.’”
It is important to note that this ruling pertains specifically to non-generative AI technologies, unlike many contemporary systems that utilize generative AI, such as large language models (LLMs). Ross Intelligence ceased operations in 2021, labeling the lawsuit as “spurious” but stating it lacked sufficient funding to continue amidst ongoing legal challenges.
As highlighted by Wired, Judge Bibas criticized Ross’s arguments, stating in his decision that “none of Ross’s possible defenses holds water” regarding the copyright infringement claims. He notably rejected Ross’s fair use argument, emphasizing the detrimental impact of its actions on the market value of Thomson Reuters’ original works, describing Ross as a direct competitor.
The foundation of Thomson Reuters’ lawsuit revolved around Ross’s usage of its Westlaw search engine, which not only indexes vast amounts of publicly available legal material but also incorporates proprietary content. Westlaw features unique elements such as headnotes—brief summaries crafted by legal experts intended to enhance the attractiveness of the subscription service for legal professionals.
In its development of a legal research platform, Ross utilized Westlaw’s annotations and headnotes, transforming them into numerical data that was subsequently used to train its AI models. The court’s ruling reveals that after Thomson Reuters denied Ross’s request to license this content, Ross sought an alternative source, purchasing 25,000 Bulk Memos from LegalEase that contained inquiries formulated by attorneys relying on Westlaw’s headnotes.
Responding to the accusations, Ross CEO Andrew Arruda argued that the Westlaw data functioned merely as “added noise” and that the company’s AI tool was designed to extract answers from legal texts independently through machine learning. Nevertheless, Judge Bibas stated that the evidence indicating actual copying was so compelling that no reasonable jury could reach a different conclusion after analyzing the similarities among the 2,830 Bulk Memo questions, headnotes, and judicial opinions.
Source
www.theverge.com