AI training on copyrighted content ruled not “fair use” by US court – how could this impact the music industry?In a landmark decision, a US district court has ruled that using copyrighted material without permission to train AI does not constitute “fair use”, a ruling that could have far-reaching implications for the music industry’s ongoing legal battles against AI companies.
The case in question saw news conglomerate Thomson Reuters, proprietor of the renowned Reuters news agency, face off against Ross Intelligence, a now-defunct legal AI startup.
The dispute centred on Ross Intelligence’s use of Thomson Reuters’ Westlaw legal research database to train its AI-powered search engine without authorisation. Ross argued that its use of copyrighted material qualified for a “fair use” exemption – an argument strikingly reminiscent of those made by AI firms facing lawsuits from the music industry.

READ MORE: So…How did Apple Music become a hero of streaming services, and Spotify a supervillain?

On 11 February, Judge Stephanos Bibas of the US District Court for the District of Delaware rejected Ross’s fair use defence, overturning a 2023 decision that had previously sent the case to a jury trial.
In his ruling, Judge Bibas noted that “originality is central to copyright.” And per court documents, Thomson Reuters’ headnotes (“A headnote is a short, key point of law chiseled out of a lengthy judicial opinion”) were determined to “have original value as individual works.”
“More than that, each headnote is an individual, copyrightable work,” the judge noted. “That became clear to me once I analogised the lawyer’s editorial judgment to that of a sculptor. A block of raw marble, like a judicial opinion, is not copyrightable. Yet a sculptor creates a sculpture by choosing what to cut away and what to leave in place. That sculpture is copyrightable.”
The court also ruled that Ross’s use of Thomson Reuters’ data was “not transformative” because, “it does not have a ‘further purpose or different character’ from Thomson Reuters’”
“Even taking all facts in favor of Ross, it meant to compete with Westlaw by developing a market substitute. And it does not matter whether Thomson Reuters has used the data to train its own legal search tools; the effect on a potential market for AI training data is enough.”
Ross’s use was “commercial”, which undermined its fair use defense.
While the ruling appears to favour copyright holders, Judge Bibas stressed that his decision specifically addressed non-generative AI, stating, “Because the AI landscape is changing rapidly, I note for readers that only non-generative AI is before me today”.
This distinction opens the door for future debates, particularly in cases involving generative AI tools, such as those used to create music, like Suno or Udio. These tools don’t just analyse and categorise data — they generate new content, and are likely to be treated differently under copyright law.
Regardless, the Thomson Reuters versus Ross case marks a crucial moment in the ongoing legal conversation surrounding AI and copyright. It signals that courts are becoming more vigilant in scrutinising the relationship between AI and copyrighted material, which would hopefully pave the way for more cases in which the music industry — among others — fights to retain control over their intellectual property.
The post AI training on copyrighted content ruled not “fair use” by US court – how could this impact the music industry? appeared first on MusicTech.