CNN sues Perplexity over verbatim article copies

CNN filed a copyright and trademark infringement lawsuit against Perplexity in the U.S. District Court for the Southern District of New York, alleging the company "unlawfully crawls, scrapes, copies, and distributes" CNN content, including paywalled stories, and reproduces "verbatim" portions of CNN articles in its answer outputs (CNN; NPR; The Verge; CNET). The complaint asserts Perplexity copied more than 17,000 items - articles, videos, and images - and says earlier licensing negotiations failed before CNN demanded the company stop using its content. Perplexity's chief communications officer, Jesse Dwyer, responded, "You can't copyright facts," a line the company has used in prior disputes (CNET).

Editorial analysis - technical context: The dispute centers on two practices common to AI answer engines - automated crawling and scraping of publisher sites, and generation of short answers that can contain verbatim or closely paraphrased source text. Coverage also cites claims that Perplexity accessed paywalled content and at times attributed inaccurate or "hallucinated" passages to publishers, which feeds both the copyright and the trademark arguments (CNET; The Verge; Engadget).

CNN's action - reportedly the first by a television network - joins a wave of suits against Perplexity. As of late May 2026, at least nine organizations had active cases, including the New York Times, News Corp and Dow Jones, the New York Post, the Chicago Tribune, Encyclopedia Britannica, Merriam-Webster, Reddit, and Japan's Yomiuri Shimbun (NPR; Press Gazette). Other publishers - among them Time, Gannett, and Le Monde - have signed licensing deals instead. Courts have not settled how copyright law applies to training data and model outputs, so the case sits in an unsettled landscape that will shape licensing, liability, and product design for answer engines.

Editorial analysis: Track whether CNN seeks and wins injunctive relief or damages early; whether the court treats reproductions in answer outputs as actionable copying versus unprotected factual extraction; any settlement or licensing terms that emerge; and precedents on scraping paywalled content. The complaint and the filing by CNN's firm, Rothwell Figg, are the primary documents for precise claims and counts.