Federal Appeals Court Deals Blow to Use of Copyrighted Works to Train AI
The 3rd Circuit Court of Appeals handed down a ruling Tuesday (Sept. 29) that Ross Intelligence’s use of copyrighted material to train its legal artificial intelligence search engine was not fair use and affirmed a lower court’s grant of summary judgment to Thomson Reuters. The ruling is the first by a federal appeals court, setting […] The post Federal Appeals Court Deals Blow to Use of…
On September 29, the 3rd Circuit Court of Appeals issued a ruling that poses a significant challenge to Ross Intelligence's practice of utilizing copyrighted material for training its artificial intelligence legal search engine. This decision marks a landmark precedent in the ongoing legal disputes concerning AI and copyright.
The court's attention was drawn to Ross Intelligence's unauthorized use of headnotes - concise legal annotations typically found in Thomson-owned Westlaw's legal research platform - to train its AI-powered search engine, which was a direct competitor to Westlaw's service. The court's 32-page opinion ruled that Ross Intelligence's use of the headnotes led to detrimental consequences for Thomson Reuters, hindering the latter's ability to promote its service.
The Copyright Act outlines four factors that courts must consider when determining whether the unauthorized use of copyrighted material constitutes non-infringing fair use. Among these factors, the effect of the use on the potential market or value of the copyrighted work is often the most decisive. In this case, the court emphasized that the headnotes served as a crucial feature for Thomson Reuters, attracting users to purchase Westlaw subscriptions.
By copying the headnotes and incorporating them into its own competing platform, Ross Intelligence appropriated valuable content, diminishing its worth as an attractive offering for Thomson Reuters' legal research platform.
The appeals court's ruling, while a setback for proponents of fair use in the AI-copyright debate, distinguishes itself from earlier district court rulings that found the use of copyrighted works for AI model training to be fair use. Notably, Ross Intelligence's search engine was not generative; it merely provided passages directly extracted from judicial opinions in response to user queries.
The court recognized that Ross's use of Thomson Reuters' headnotes was not highly transformative, as both companies aimed to create and optimize legal research platforms to aid users in finding relevant legal materials. This lack of transformation weighed against a finding of fair use, the opinion stated.
Furthermore, the 3rd Circuit rejected Ross Intelligence's attempt to argue that the case transcended copyright concerns, emphasizing that the issue fundamentally revolves around copyright infringement. A spokesperson for Thomson Reuters, owned by the news agency Reuters, expressed the company's unwavering belief in the importance of respecting copyright to foster innovation while safeguarding intellectual property that drives AI solutions.
Written by urgent.news from PYMNTS's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
