In May 2020, Thomson Reuters, the media and technology conglomerate, has set up a small legal AI startup called Ross Intelligence to use U.S. copyright law by duplicating material from Westlaw, Thomson Reuters’ legal research platform. A lawsuit was filed for violating the law. As the pandemic raged on, the case was largely unknown outside the small world of geeks obsessed with copyright rules. But the lawsuit, filed more than two years before the generative AI boom began, turned out to be a preemptive strike in a larger war between content publishers and artificial intelligence companies currently being waged in courtrooms across the country. It is now clear. The results could create, destroy, or reshape entire information ecosystems and the AI industry, and in doing so, impact nearly everyone on the internet.
Dozens of other copyright lawsuits against AI companies have been rapidly filed over the past two years. Plaintiffs include independent authors and visual artists such as Sarah Silverman and Ta Nehisi-Coates, media companies such as the New York Times, and music industry giants such as Universal Music Group. The various rights holders claim that AI companies are using the results to train powerful, often highly profitable AI models in a manner that amounts to theft. AI companies often rely on the so-called “fair use” doctrine to protect themselves, meaning that the construction of AI tools can be protected by copyright without obtaining consent from or paying compensation to the rights holders. It argues that these should be considered situations in which it is legal to use such material. (Widely accepted examples of fair use include parody, news reporting, and academic research.) Nearly every major generative AI company has been involved in this legal battle, including OpenAI, Meta, Microsoft, Google, Anthropic, and Nvidia. is caught up in.
WIRED is closely monitoring how these lawsuits develop. We track which companies and rights holders are involved, where the lawsuit was filed, what they’re alleging, and everything else you need to know to help you stay on top of the situation. I created a visualization to do this.
The first case, Thomson Reuters v. Ross Intelligence, is still pending in court. The trial, originally scheduled for earlier this year, has been postponed indefinitely, and it is unclear when it will end, with legal costs already putting Ross out of business. Other cases, such as the high-profile case brought by the New York Times against OpenAI and Microsoft, are currently in contentious discovery periods, during which what information the parties will be required to submit. are being discussed.
