CNN Files First-of-Kind AI Copyright Suit Against Perplexity, Tests News Distribution Theory

CNN Fires First Shot in TV News Copyright War

CNN filed suit against AI search engine Perplexity on May 28, 2026, in the Southern District of New York, becoming the first television network to bring copyright claims against a generative AI company. The case tests untested legal theories about how AI systems distribute news content versus merely training on it.

The complaint alleges Perplexity "unlawfully copied and distributed" CNN's content after failed licensing negotiations in 2025. This isn't your standard training data dispute. CNN argues Perplexity's search responses constitute direct distribution of copyrighted material to users, bypassing the fair use defences that have protected other AI companies in training-focused litigation.

"CNN's lawsuit stands for the proposition that Perplexity, a company valued at tens of billions of dollars, should not be able to steal from entities that create the original content," the network said in a statement.

Distribution vs. Training: The Legal Distinction

Most AI copyright suits challenge the ingestion of training data. CNN's theory targets the output stage. When Perplexity generates responses using CNN content, the network argues, it's not transforming that content for training purposes but distributing it as a substitute news product.

This matters because courts have shown more sympathy for training-based fair use claims than distribution theories. Judge William Alsup granted summary judgment to Meta last year on training claims, writing that AI model development was "exceedingly transformative." But distribution claims face different legal standards.

Perplexity's response suggests it recognises the distinction. "You can't copyright facts," chief communications officer Jesse Dwyer told CNN, focusing on the information rather than its expression. That's exactly the wrong answer if CNN's distribution theory gains traction.

Failed Negotiations Create Stronger Infringement Case

The complaint reveals CNN and Perplexity attempted to negotiate a content licensing deal in 2025 but "did not agree on terms." This strengthens CNN's position considerably. Unlike cases where AI companies scraped content without any publisher contact, Perplexity knew CNN's position on unauthorised use.

"Before and after Perplexity's negotiations with CNN, Perplexity knew that it was not permitted to access CNN's content," the lawsuit states. That language suggests wilful infringement, which can trigger enhanced statutory damages under 17 U.S.C. § 504(c).

The failed licensing talks also undercut fair use defences. Courts consider the effect on potential markets when evaluating fair use under the fourth factor. If CNN was willing to licence and Perplexity chose to proceed without permission, that weighs against transformation arguments.

Television Content Presents Unique Challenges

CNN joins a crowded field of media plaintiffs challenging AI content use. The New York Times, Chicago Tribune, and other publishers have filed similar claims. But television content presents different copyright challenges than print media.

News broadcasts involve multiple copyrighted elements: scripts, video footage, graphics, and potentially music. CNN's complaint doesn't specify which elements Perplexity allegedly infringed, but video content typically receives stronger protection than factual news text.

The case highlights the AI industry's licensing divide. While some companies like OpenAI have struck content deals with publishers, others continue operating under fair use theories. CNN's lawsuit suggests that approach carries real litigation risk.

What Happens Next

Perplexity will likely move to dismiss on several grounds: fair use, failure to state a claim, and lack of substantial similarity between inputs and outputs. The company may also challenge CNN's damages theory, since proving market harm for individual news articles remains difficult.

But CNN's distribution argument could survive initial motions if the court finds Perplexity's responses constitute more than fair use commentary. Recent rulings in Advance Local Media v. Cohere suggest judges are willing to let "substitutive summary" theories proceed to discovery.

The case will test whether AI companies can continue treating news content as free training data when their business models directly compete with news distribution. For television networks sitting on vast archives of copyrighted content, CNN's approach offers a litigation roadmap that doesn't depend on proving training data harm.

Perplexity faces exposure to statutory damages of up to $150,000 per infringed work. If CNN can prove access to thousands of articles, the math becomes existential quickly. That's the same dynamic that drove Anthropic's $1.5 billion settlement with book authors last year.

The next procedural step is Perplexity's response to the complaint, due within 21 days of service. The real test comes when the court rules on fair use as a matter of law versus sending distribution questions to a jury.