Court upholds Thomson Reuters in AI and copyright case

A US court has ruled in favour of Thomson Reuters in a high-profile case over artificial intelligence (AI) and copyright. The judge rejected the “fair use” argument and found that copying content without permission was illegal. The defendant in the case was AI startup Ross Intelligence, which was trying to build a competing product based on Thomson Reuters data.

Image source: Conny Schneider / Unsplash

As PCMag writes, the lawsuit was filed back in 2020. Thomson Reuters accused Ross Intelligence of illegally using materials from the paid legal platform Westlaw. In particular, the startup copied closed summaries of provisions of a particular law in specific court cases, which were compiled by Westlaw editors. Judge Stephanos Bibas ruled that such summaries are protected by copyright, and their use without permission violates the law.

The AI ​​startup tried to defend itself by arguing that its actions were “innocent infringement,” but the court did not accept that argument. The judge also considered the standard fair use criteria (whether the use is commercial or non-commercial), the nature of the protected material, and the impact on the market. In the end, two factors favored Ross Intelligence and two favored Thomson Reuters.

The decisive factor was the damage to the market. The court ruled that Ross Intelligence used the content to create a competing product, thereby violating copyright. At the same time, the judge emphasized that this was not about generative AI, but about direct copying of texts without permission.

The ruling could set an important precedent, especially given the slew of lawsuits against AI developers. For example, The New York Times is suing OpenAI and Microsoft, accusing them of illegally using journalistic materials. Perplexity has been sued by the American media holding News Corp, and several other major Canadian media outlets are also suing OpenAI.

Interestingly, some companies have opted to enter into commercial agreements rather than litigate. For example, media holding Vox Media and The Atlantic magazine signed a deal with OpenAI, and news site Axios signed a deal with Meta✴. These agreements will allow the companies to control how their materials are used in AI models.