Login | April 26, 2025
Westlaw gets partial win in AI scraping copyright case: Analysis
RICHARD WEINER
Technology for Lawyers
Published: April 11, 2025
If you hadn’t seen the news (and if you are in IP, you surely did), a judge has ruled in favor of Thomson Reuters and Westlaw in an AI scraping copyright case against some company called Ross Intelligence.
The judge rejected Ross’s Fair Use argument in Thomson Reuters Enterprise Centre GmbH v. Ross Intelligence Inc., No. 1:20-CV-613-SB (D. Del. Feb. 11, 2025) (link to decision below).
The case, a first in what will be a long line of such cases, involved Ross engaging a third party to scrape Westlaw to eventually train its AI to create paragraph-long summaries of cases for Ross’s customers. Or something very similar to that.
The judge said “No, Ross! You can’t do that! It’s a copyright violation!”
Anyway, the implications are potentially enormous, even though this is a partial ruling with little-to-no precedent.
Analyst Chris Hanslik at BoyarMiller breaks the decision down into several “implications for business.” Here is a combo of that one with my little contribution.
The first and most obvious is copyright protection and AI training data. The decision straightforwardly suggests that AI trainers have to be more diligent in ensuring that their training doesn’t infringe on the copyrights of the people whose data is being scraped. Do the due diligence (fer cryin’ out loud). It isn’t that hard.
Fair Use. The case now establishes some limits on the Fair Use argument in data scraping copyrighted materials. While not exactly precedent, courts in other ongoing AI data scraping cases like the one between OpenAI and the NYT will likely take a look at this court’s analysis.
Data privacy and compliance risks. The gloves might be off at the new SEC, for instance, but the rest of the world still has laws like GDPR, and state laws like the CCPA may come into play here. So now that AI scrapers have to pay more attention to laws and things like that, they could also look at laws beyond just copyright.
Competitiveness. “Clean” scraping could become a thing now as a competitive advantage. And then, of course, we would have “clean washing.” Cause you know that will happen.
Future regulation. Maybe, but the “new” SEC doesn’t look like they will be regulating AI much, especially if there’s money involved. There has never been the will or skill to pass any national data protection laws beyond copyright. Maybe state laws can control the coming out-of-control AI blitz. But we shall see.