CNN Takes Perplexity to Court Over 17,000 Stolen Assets — And It Won’t Be the Last

Perplexity AI is no stranger to courtrooms, but the lawsuit filed against it by CNN on May 28, 2026 marks a new chapter in the escalating war between artificial intelligence companies and the news organisations whose work they depend on. It is the sixth active copyright lawsuit the AI search startup now faces from major publishers and the first time a television broadcaster has entered the fray.

Filed in a New York federal court, CNN’s 54-page complaint alleges that Perplexity systematically scraped more than 17,000 pieces of the network’s content — stories, videos, images, and other protected works from CNN’s own platforms and third-party websites. That material, CNN claims, was used to fuel Perplexity’s large language models and power the AI-generated answers it delivers to users in real time.

The trademark dimension of the complaint is equally pointed. CNN argues that Perplexity violated its trademark by presenting CNN’s journalism under its own brand without linking back to CNN’s platforms a practice that deprives CNN of the traffic, advertising revenue, and subscriber conversions that are essential to its business. According to NPR’s reporting cited in the complaint, Perplexity has on occasion generated false information attributed to CNN which the filing describes as not merely inaccurate but actively damaging to the network’s trademark.

CNN’s statement filed alongside the complaint was unambiguous: Perplexity, a company valued at tens of billions of dollars, should not be able to steal from entities that create the original content it exploits. The public relies on high-quality journalism reported by human beings to understand their world, which is frequently dangerous and expensive to produce, and commercial operators can and must pay to use it.

A Failed Negotiation, Then Continued Scraping

CNN attempted to reach a licensing agreement with Perplexity last year but was unable to settle on terms governing how the company could use CNN’s content. After negotiations broke down, CNN blocked Perplexity’s bot from accessing its material — yet the company allegedly continued scraping CNN’s content regardless. That timeline — failed talks, an explicit technical block, and ongoing access forms a critical part of CNN’s legal argument that Perplexity’s conduct was not inadvertent.

The Arguments on Both Sides

The legal tension at the centre of all six publisher lawsuits comes down to two competing positions that the courts have not yet fully resolved.

Publishers argue that journalism is expensive, dangerous, and economically sustainable only when the organisations producing it can monetise their work. When AI systems scrape, reproduce, and summarise that journalism without directing users back to the original source, they destroy the traffic and subscription model that makes the reporting possible in the first place. Licensing, they say, is the only workable solution.

Perplexity’s defence rests on a more familiar intellectual property principle. The company has argued in multiple legal responses that attempts to stop AI search by monopolising facts will fail on bedrock principles of intellectual property law that have consistently permitted innovative technologies like Perplexity to exist. In the Dow Jones case, Perplexity called the adversarial posture fundamentally shortsighted, unnecessary, and self-defeating. A company spokesperson responded to the CNN lawsuit with a similarly blunt three-word position: “You can’t copyright facts.”

CNN’s complaint, however, is designed to cut through that argument. The complaint claims that Perplexity copied not just the facts but the specific creative ways those facts were presented the writing, reporting, editorial choices, video production, and photography that constitute the protected work. That framing has considerably stronger support in existing copyright law.

A Divided Industry

Perplexity finds itself navigating an industry that has split sharply on how to respond to AI companies. While it fights lawsuits from the New York Times, Dow Jones and News Corp, the Chicago Tribune, and Encyclopaedia Britannica and Merriam-Webster, it has also struck revenue-sharing licensing deals with TIME, Fortune, Gannett, Le Monde, Der Spiegel, the Los Angeles Times, and The Independent. Perplexity’s licensing model offers publishers a revenue share exceeding 10%, which differs from OpenAI’s approach of paying a flat annual fee of between $1 million and $5 million to most partners with News Corp’s OpenAI deal reportedly valued at more than $250 million over five years.

The commercial costs of Perplexity’s legal exposure are already showing. Snap terminated its $400 million search deal with Perplexity just six months after beginning tests.

Why CNN’s Case Could Carry Extra Weight

CNN’s complaint is distinctive because it involves a television network and alleges copyright violation across multiple media types including video and images whereas earlier cases brought by the New York Times and Dow Jones dealt primarily with text. The breadth of CNN’s claim could extend the legal implications to broadcasters and media companies that have not yet entered litigation.

The timing adds further context. The full enforcement provisions of the EU AI Act are set to take effect on August 2, 2026, adding regulatory pressure to an already volatile legal environment. And the broader litigation landscape has shifted meaningfully: the $1.5 billion settlement in the Bartz v. Anthropic class action involving approximately 120,000 authors is the largest copyright settlement in US history and signals that AI companies can face substantial financial exposure for content-related claims, even while exact legal standards remain contested.

What is already clear is that the divide between AI companies that pay for journalism and those that don’t is hardening. CNN’s lawsuit is the most recent instalment in that reckoning — and, by most indications, far from the last.