CNN has sued Perplexity AI in federal court in New York, accusing the AI search company of unlawfully copying and redistributing CNN journalism through its answer engine and related products.
The complaint, filed in the U.S. District Court for the Southern District of New York, alleges that Perplexity used thousands of CNN stories, videos and images without permission and generated responses that reproduced or closely mirrored CNN reporting. CNN is seeking damages and an order blocking further alleged infringement.
The case is part of a wider publisher challenge to AI search and answer engines. The dispute is not only about whether copyrighted material can be used to train AI systems. It is also about whether an AI product can act as a substitute for the original article by giving users the substance of reporting without sending them to the publisher that paid to produce it.
CNN also alleges that Perplexity bypassed technical measures intended to restrict automated access to its content. Perplexity has rejected publisher claims in similar disputes, arguing that facts cannot be copyrighted. That defense may be harder to sustain where a complaint alleges near-verbatim reproduction or distribution of protected expression rather than the reuse of facts alone.
The lawsuit also includes trademark-related claims, with CNN alleging that Perplexity created confusion around whether CNN content was included in a “premium news bundle” or otherwise affiliated with the service.
For media companies, the operational issue is becoming clearer. AI licensing is not just a future training-data negotiation. It is now tied to search, summarization, paywall leakage, attribution, traffic substitution and the packaging of publisher brands inside third-party AI products.
The case joins a growing list of legal actions against Perplexity from publishers and content owners, including The New York Times, Dow Jones, the New York Post, Encyclopedia Britannica, Merriam-Webster and Reddit. Some publishers have chosen licensing deals with AI companies; others are testing whether courts will force a different economic model.
Rights strategy for news and factual content is moving from the archive to the interface. Publishers are asking not only who can train on their work, but who can answer with it, summarize it, sell access around it and potentially replace the click.