CNN sues Perplexity

Hey guys, Mr. Technology here. The first major TV network just took Perplexity to federal court over AI copyright — and the lawsuit is more specific and more technical than the New York Times case from 18 months ago. Every AI search/answer product needs to read this complaint.

What You Need to Know: CNN sued Perplexity in the Southern District of New York on May 28, 2026, alleging the AI search company scraped more than 17,000 CNN stories, photos, and videos to train and serve its products. The 54-page complaint is the first copyright/trademark suit by a TV network against an AI company, and it comes after CNN tried and failed to license a deal — Perplexity's terse response: "You can't copyright facts."

Why It Matters

• The "facts vs. expression" defense is going to be tested in court. Perplexity's four-word statement — "You can't copyright facts" — is the strongest First Amendment argument available, and it might win on the copyright count. But CNN's complaint is built on trademark and unfair competition theories that don't depend on the facts-vs-expression line.
• Scraping-after-being-blocked is the new smoking gun. The complaint says CNN explicitly blocked Perplexity's scraper bot and Perplexity kept going. If you run an AI search product, your robots.txt discipline and proxy management are now your single biggest legal risk.
• "Comet Plus" branding is the trademark wedge. CNN's strongest claim is that Perplexity marketed a "Comet Plus" tier as if it had a CNN content relationship when it didn't. The trademark and false-advertising claims may survive even if the copyright claims get narrowed.
• The pattern is accelerating. New York Times, Dow Jones (Wall Street Journal), New York Post have all sued. Time and USA Today Co. (formerly Gannett) struck deals. Meta signed CNN in December 2025. The negotiation-vs-litigation fork is happening live.

What Actually Happened

The Lawsuit: 54 Pages, Two Counts, One Newsroom

CNN filed its complaint in the U.S. District Court for the Southern District of New York on Thursday, May 28, 2026. The 54-page filing alleges Perplexity "scraped more than 17,000 CNN stories, photos, videos and other content" and used that material both to train its products and to serve answers in response to user queries. (Variety, Reuters)

The case is the network's first AI-copyright lawsuit and is believed to be the first of its kind brought by a U.S. television network. CNN's public statement, issued through a spokesperson, framed the suit as forced by Perplexity's refusal to negotiate: "The public rely on high-quality news journalism reported by human beings to understand their world, which is frequently dangerous and expensive to produce. Commercial operators can and must pay to make use of it. We prefer that they do so through sensible licensing arrangements, but if they refuse to do that as Perplexity has so far refused to do, they will have to pay through legal damages. There is no free option."

The Trademark Claim Is the Strongest Part of the Case

Most AI copyright cases live or die on the "facts vs. expression" doctrine — and the New York Times' 2023 suit against OpenAI has struggled to overcome the defense that news facts themselves are not copyrightable. CNN's lawyers clearly anticipated that fight and built a parallel case on trademark and unfair competition. (Variety)

The complaint highlights Perplexity's marketing of a "Comet Plus" subscription tier that the network alleges falsely advertised a content relationship with CNN. The complaint reads: "Perplexity's use infringes CNN's exclusive rights in its federally registered trademarks, and has caused and is likely to cause confusion, mistake, or deception as to whether the articles Perplexity provides are associated or affiliated with, or are sponsored, endorsed, or approved by CNN."

Trademark claims are harder for AI companies to dismiss on First Amendment grounds because they don't require proving the copying of a creative work — they require proving consumer confusion about source affiliation. CNN is asking the court for statutory damages and an injunction blocking Perplexity from using its content.

The Scraping-After-Block Is the Smoking Gun

The complaint contains an allegation that should make every AI search operator sit up straight: CNN says it spoke to Perplexity about a licensing deal in 2025, negotiations failed, and the network then blocked Perplexity's scraping bot from accessing its content. Perplexity allegedly continued accessing CNN's content anyway. (Variety)

That sequence — block, then continue — is the part that will hurt Perplexity most in court. Most AI-search copyright cases turn on whether the scraping was licensed; CNN has positioned this one to turn on whether scraping continued after an explicit stop. If the network can produce server logs showing requests from Perplexity's known IP ranges after the block, the case becomes much harder to settle and much easier to win at trial.

Perplexity's Response: Four Words

Asked for comment, a Perplexity spokesperson emailed Variety the four-word statement: "You can't copyright facts." The line is a direct invocation of the long-standing copyright principle that raw facts — names, dates, statistics — are not protectable expression. It's the same defense the company used in 2024 against Dow Jones, when it called that lawsuit "fundamentally shortsighted, unnecessary and self-defeating." (Variety)

Whether the four-word defense wins on the copyright count is genuinely uncertain. But it does not address the trademark and unfair-competition claims, and it doesn't address the post-block scraping. Expect Perplexity to file an early motion to dismiss focused on the copyright counts; expect CNN to lean on the trademark counts as the case's structural backbone.

The Wider Pattern: Lawsuits and Licensing Deals

CNN's filing is the latest in a growing list of publisher actions against Perplexity. The New York Times filed its own lawsuit in late 2025, and Dow Jones (publisher of the Wall Street Journal) and the New York Post both sued in 2024. Meanwhile, Time and USA Today Co. (formerly Gannett) struck content deals with the company. Meta signed its own CNN content-licensing deal in December 2025. (Variety)

The split is telling. The publishers that are suing are the ones whose content is most differentiated — and most expensive to produce. The ones that took deals are the ones whose content is more commodified. Perplexity is now litigating with the high end and licensing with the middle, and the CNN case is going to be the test of whether that strategy survives the courts.

The Take

CNN's complaint is the best-constructed AI copyright case I've read this year, and the trademark wedge is the move. Most of the 2024-era AI copyright cases were built on shaky fair-use arguments and weakly-defined damages. CNN's lawyers built a case that doesn't need the copyright count to win — they have trademark, false advertising, and post-block scraping, each of which is a clean cause of action with cleaner remedies.

If you're building an AI search, AI agent, or AI answer product, here is what you should do this week:

1. Audit your robots.txt compliance and proxy management. The post-block scraping allegation is the single most damaging factual theory in this case. If you don't have clean, auditable logs showing you stopped when told to stop, you're exposed. 2. Audit your marketing language. If you sell a "premium" tier that includes named-publisher content, you need a paper trail showing the relationship. "Comet Plus" is the lesson. 3. Map your content sources by legal exposure. Differentiated, expensive-to-produce content is a litigation risk. Commodified, broadly-licensed content is a deal-making opportunity. Sort your dependencies accordingly. 4. Track the case docket closely. A ruling on trademark-vs-fair-use for AI search summaries would reset every AI licensing deal on the planet.

Quick Summary

CNN sued Perplexity for scraping 17,000+ pieces of content after explicitly blocking its bot — a case built more on trademark and false advertising than on copyright, and one Perplexity will struggle to dismiss on the four-word "facts aren't copyrightable" defense. Every AI search product should audit its robots.txt compliance and marketing language this week.

Sources

• CNN Sues Perplexity, Alleging AI Company Infringed Its Copyrights (Variety)
• CNN sues Perplexity over alleged AI copyright theft (CNN Business)
• CNN files lawsuit against Perplexity alleging unlawful content distribution (Reuters)
• Perplexity's four-word response to CNN's lawsuit (Brian Stelter on X)
• CNN v. Perplexity complaint (PDF via Variety)