News Corp. CEO Robert Thomson didn’t mince words when addressing AI companies sucking up public data without regards to who owns it: “If you have received stolen goods, we intend to pursue you relentlessly.”
The comments, which came at the top of the company’s fiscal first-quarter conference call, signals how important the issue of AI and intellectual property is to media companies. It comes amid an industry-wide debate about what kind of content is being used to train AI models, which then in turn can spit out images and videos that reference well-known IP.
““You may not have done the actual stealing, but receiving stolen property is an offense in legal jurisdictions around the world, content, crime does not and will not pay,” he added.
Thomson referenced that there’s a lot of “suing and wooing” going on, and teased that News Corp. was working on a number of licensing deals with AI companies in which its IP could be used to “ethically” train large language models. He said those deals would benefit the company financially, but didn’t offer specifics.
While much of the controversy lately has been focused on OpenAI and its Sora app, Thomson praised the company for its respect for IP. Last year, News Corp. and OpenAI signed a multi-year deal to bring news content to OpenAI’s models.
Thomson also praised Anthropic’s $1.5 billion settlement that compensates authors for books that the startup illegally obtained and used to train its AI models.
