- U.S. Circuit Judge rules in favor of Thomson Reuters against AI company Ross Intelligence in landmark copyright case.
- Ross Intelligence obtained approximately 25,000 legal documents through LegalEase to train their AI system without proper authorization.
- The court identified 2,243 instances of copyright infringement from Westlaw’s headnotes.
- Case sets precedent for ongoing AI training data disputes, including actions by prominent authors and The New York Times.
- Judge reversed previous fair use opinion, placing burden of proof on Ross Intelligence.
In a landmark ruling that could reshape AI training practices, Thomson Reuters secured a victory against legal AI developer Ross Intelligence over unauthorized use of copyrighted content from its Westlaw platform, according to court documents.
The dispute originated in May 2020 when Thomson Reuters alleged that Ross Intelligence circumvented licensing requirements by acquiring training data through LegalEase Solutions, a Michigan-based legal research provider. The court found that approximately 25,000 "Bulk Memos" containing legal queries and responses were derived from Westlaw’s proprietary headnotes.
Judge Stephanos Bibas identified 2,243 instances of copyright infringement, dismissing Ross Intelligence’s defenses of innocent infringement and copyright misuse. "When a Bulk Memo question resembles a headnote more than the original judicial opinion, it strongly suggests actual copying," the judge stated in his ruling.
This case emerges amid growing tensions between content creators and AI developers. Notable authors including George RR Martin, John Grisham, and Michael Connelly have initiated legal actions against AI companies. In December 2023, The New York Times filed suit against OpenAI over similar copyright concerns.
The ruling’s significance extends beyond immediate parties, potentially establishing precedent for how AI companies can legally acquire training data. LegalEase’s involvement revealed systematic efforts to rewrite Westlaw content, with internal guides specifically instructing against direct copying of headnotes – a practice the court ultimately found insufficient to avoid copyright infringement.
"Smart man knows when he is right; a wise man knows when he is wrong," Judge Bibas noted while revising his earlier opinion on fair use, placing the burden of proof squarely on Ross Intelligence. This revision could influence how courts approach fair use claims in future AI training data disputes.
The case underscores growing legal scrutiny over AI training practices and may prompt AI developers to seek explicit permissions for training data usage, potentially increasing development costs and timeline considerations for future AI projects.
✅ Follow BITNEWSBOT on Telegram, Facebook, LinkedIn, X.com, and Google News for instant updates.
Previous Articles:
- Fireblocks Expands Support to Sony’s New Soneium Blockchain Network
- Franklin Templeton Signals Potential Spot Solana ETF Launch with Delaware Trust Filing
- SEC Crypto Task Force Chief: Meme Coins Likely Outside Regulatory Scope
- Uniswap Launches New Layer-2 Blockchain with One-Second Block Times
- Uniswap Labs Launches UniChain L2 Network, Targets MEV Protection with Flashbots TEE Integration
