Thomson Reuters wins early AI copyright case
Judge rules against fair use claim in lawsuit over Westlaw content used by Ross Intelligence
Thomson Reuters has won an early, landmark court battle over AI and copyright in its case against legal AI startup Ross Intelligence.
In 2020, Thomson Reuters sued Ross Intelligence for copyright infringement, claiming that Ross used Westlaw’s editorial content—such as headnotes and legal annotations created by human editors—as training data for its AI tool without permission. Ross argued that its use of the content was "fair use."
US District Court Judge Stephanos Bibas issued a summary judgment in favor of Thomson Reuters, rejecting all of Ross’s defenses. The judge stated, "None of Ross’s possible defenses holds water," noting that the copying was so evident that no reasonable jury could find otherwise. A key factor in the decision was that Ross’s use of the material directly competed with Westlaw by creating a market substitute, thereby undermining the market value of the original content.
Ross Intelligence had already shut down in 2021 due to the high cost of litigation, and this ruling sets an important precedent as similar lawsuits are underway against other AI companies, such as OpenAI and Microsoft. Legal experts warn that if this decision is followed in other cases, it could significantly weaken the fair use defenses that many generative AI companies rely on.
This decision represents one of the first major legal tests of AI copyright issues and may have broad implications for how AI companies train their models using copyrighted material in the future.
Sources: The Verge, Wired, Reuters.
