CNN grants Google enhanced AI content rights in multi-year deal

CNN announced a strategic, multi-year partnership with Google on Oct. 15 that gives the technology company enhanced rights to CNN journalism for experimental features in Google products. The companies will explore how generative AI can help audiences discover and engage with CNN’s reporting. The Skydance-owned news network said the arrangement is intended to open new ways for audiences to reach its journalism while compensating CNN for the delivery and use of its intellectual property in connection with AI products. The agreement places intellectual-property safeguards alongside CNN’s stated effort to pursue AI opportunities through commercial partnerships and discussions with responsible industry players. Sam Felix, CNN’s senior vice president for strategic partnerships and business development, said Google’s generative-AI innovation had changed how people discover information and content. He said Google’s tools would help CNN explore new ways of finding and engaging audiences, while the partnership would provide payment for use of CNN’s journalism in AI products. The agreement arrives as an AI content-licensing market takes shape. Brookings said AI systems are increasingly using journalistic information without necessarily directing users back to publishers, potentially weakening advertising revenue and subscriber conversion. It described bilateral licensing agreements as confidential arrangements between major AI companies and selected publishers, generally those with national or global brands. That model does not necessarily protect publishers from reduced referrals. The initial click-through advantage for publishers with direct AI agreements had essentially disappeared by the fourth quarter of 2025, when click-through rates from AI systems had fallen six-fold. Publishers with and without direct deals both lost AI-driven referrals, although the proportional decline was smaller for those without agreements. Voluntary licensing can create fragmented and selective coverage, yielding incomplete or biased datasets. Terms, pricing precedents and governance norms being established now may become difficult to alter once they are normalized.