Thomson Reuters won an early court battle over whether artificial intelligence (AI) programs can be trained on copyrighted materials.
The media company filed a lawsuit in 2020 against Los Intelligence, a now-retired legal research company. In it, Thomson Reuters claims that the company used its own legal platform Westlaw to train AI models without permission.
In his decision, Judge Stefanos Vivas confirmed that Ross Intelligence is not permitted under US copyright law, known as the “fair use doctrine.”
Related
The “fair use” doctrine of US law limits the use of copyrighted material, such as education, research and conversion of copyrighted works into different things.
“We are pleased that the court has granted a summary judgment in our favour,” according to the next statement from Thomson Reuters to Euroneus.
“Copying content was not “fair use.”
Ross Intelligence did not immediately respond to requests for comment from Euronows Next.
Related
Thomson Reuters’ victory comes as more and more lawsuits have been filed by authors, visual artists and music labels against the developers of AI models over similar issues.
Linking each of these cases is that high-tech companies consume a horde of human writing, train AI chatbots to obtain permission or compensate those who wrote the original work. The claim is that they created texts that are human-like.
