Perplexity AI Sued by CNN, NYT, Amazon — Inside Every Active Case

CNN sued Perplexity AI on May 28, 2026, accusing the AI search company of scraping more than 17,000 CNN stories, videos, and images to power its products without payment or permission. CNN’s complaint is the seventh major publisher lawsuit against Perplexity in under two years, following near-identical claims from The New York Times, Dow Jones, the New York Post, Encyclopedia Britannica, Merriam-Webster, and the Chicago Tribune.

The cases are filed primarily in the U.S. District Court for the Southern District of New York, alleging copyright and trademark infringement under the Copyright Act and the Lanham Act. Separately, Amazon won a preliminary injunction against Perplexity in March 2026 over unauthorized website scraping, and a privacy class action filed in April 2026 accuses the company of secretly sharing user conversations with Meta and Google. As of June 2026, all of these cases remain active.

Perplexity AI Lawsuit Timeline and Updates

October 2024 — Dow Jones and the New York Post File First

News Corp subsidiaries Dow Jones and the New York Post sued Perplexity in the Southern District of New York. The complaint accused the company of a “massive illegal copying” scheme that diverted readers and ad revenue by answering queries directly instead of linking out.

This was the opening shot. Every subsequent publisher lawsuit borrowed its core legal theory.

October 2024 — Reddit Sues Over Scraping

Reddit filed a separate New York federal lawsuit accusing Perplexity and three other companies of unlawfully scraping its platform data. Reddit’s complaint relied partly on the Digital Millennium Copyright Act, a different statutory angle than the publishers were using.

That is the pattern: each plaintiff added a slightly different legal theory to the same underlying scraping allegation.

September 2025 — Encyclopedia Britannica and Merriam-Webster Sue

Encyclopedia Britannica Inc. and Merriam-Webster Inc. filed a joint lawsuit over data scraping. Their case extended the publisher claims beyond breaking news into reference material, arguing Perplexity’s answer engine substituted for dictionary and encyclopedia lookups the same way it substituted for news searches.

No injury required a news cycle. A definition lookup was enough to trigger the same legal theory.

December 2025 — New York Times and Chicago Tribune File Parallel Suits

The New York Times and Chicago Tribune each filed separate complaints in the Southern District of New York within days of each other. The Times’s filing came more than a year after it had sent Perplexity a cease-and-desist letter.

The Times alleged Perplexity could reproduce “identical or substantially similar” text from its articles on request. It also alleged Perplexity sometimes attached fabricated “hallucinations” to the Times’s name and trademarks, creating a reputational harm distinct from the copying itself.

March 2026 — Perplexity Moves to Dismiss the Times and Tribune Suits

Perplexity filed a motion to dismiss both cases. The motion argued the Times and Tribune failed to allege supporting facts for 3,234 of the 3,240 works named in the suit, providing only 11 examples of outputs the company called substantially similar.

Perplexity also argued any infringement traced back to user prompts, not to any deliberate action the company itself took. That argument would resurface, nearly word for word, in Perplexity’s later response to CNN.

March 10, 2026 — Amazon Wins a Preliminary Injunction

U.S. District Judge Maxine Chesney, ruling in the Northern District of California, found Amazon had submitted “strong evidence” that Perplexity’s Comet browser accessed Amazon’s website “without authorization.” Amazon had sued in November 2025, alleging Perplexity’s AI shopping agent disguised itself as a normal Chrome browser session to evade detection.

Chesney noted Amazon had spent more than $5,000 and “numerous hours” building tools specifically to block Comet, and that this qualified as compensable harm. Perplexity called the suit a “bully tactic” and was granted a one-week stay to appeal.

March 31, 2026 — Privacy Class Action Filed Over Data Sharing

A proposed class action was filed in the Northern District of California alleging Perplexity embedded Meta Pixel and Google advertising trackers into its website and mobile app. The complaint claimed these trackers captured user conversations, including sensitive financial and health-related prompts, and transmitted them to Meta and Google in real time.

The case is Doe v. Perplexity AI Inc., No. 3:26-cv-02803. A Utah man identified only as John Doe is the named plaintiff, representing a proposed class that could include roughly 20 million U.S. monthly active users.

April 1, 2026 — Doe Complaint Details “Incognito” Mode Failure

The complaint’s most damaging allegation concerned Perplexity’s “Incognito” privacy setting. Plaintiffs claimed that even when users activated Incognito mode expecting anonymity, the embedded trackers continued capturing and transmitting their prompts to Meta and Google, tied to the users’ actual Facebook and Google account identifiers.

That defeats the entire purpose of an incognito feature. A privacy setting that fails silently is arguably worse than having no privacy setting advertised at all.

May 28, 2026 — CNN Files Suit, Seeking Damages and an Injunction

CNN filed its 54-page complaint in the Southern District of New York, case No. 1:26-cv-04427. The lawsuit alleged Perplexity scraped more than 17,000 CNN stories, videos, images, and other works, in some cases bypassing technical blocks CNN had specifically deployed against its crawlers.

CNN’s complaint pointed to one example where a user asked Perplexity Pro to “deeply analyze an article from a single reputable source.” The system produced text that closely tracked a CNN article registered under Copyright Registration TX 9-574-794.

May 28, 2026 — CNN Adds a Trademark Claim Tied to Comet Plus

CNN’s complaint went further than a pure copyright theory. It alleged Perplexity advertised to subscribers of its Comet Plus tier that they could access CNN’s premium content, despite no licensing relationship ever existing between the companies.

CNN had, in fact, discussed a Comet Plus arrangement in October 2025. Those talks fell apart before any deal was signed. Advertising access to content under a deal that never closed is precisely the kind of false-affiliation claim trademark law exists to address.

Why Publishers Keep Filing the Same Lawsuit Against One Company

Here is what matters: seven separate plaintiffs filed seven separate lawsuits making nearly the same argument, rather than consolidating into one case. That is not an accident of timing.

Each publisher needed to establish its own ownership of its own specific copyrighted works. A consolidated class action would have diluted the individualized proof each plaintiff needs: which articles, which registrations, which specific instances of verbatim reproduction.

Yomiuri Shimbun, the Japanese media company, also filed suit during this window, extending the litigation wave outside U.S. borders entirely. Michael Goodyear, an associate professor at New York Law School, has noted that more than 100 similar AI copyright lawsuits are now winding through federal courts nationwide, with no appellate court yet issuing a definitive ruling on whether training an AI model on copyrighted material counts as fair use.

Perplexity’s Core Legal Defense, Explained

Perplexity’s defense rests on three connected arguments, repeated with only minor variation across every publisher case.

First: facts cannot be copyrighted. Chief Communications Officer Jesse Dwyer made this argument directly to CNN, framing the dispute as an attempt to “monopolize” information rather than protect expression.

Second: compiling content into a searchable database is fair use. Perplexity’s formal answer in the Dow Jones case stated plainly that “compiling copyrighted content to create a searchable database has long been deemed fair use under US copyright law,” pointing to decades of precedent around search engine indexing.

Third: any infringing output traces back to user prompts, not to Perplexity’s own conduct. The company’s March 2026 motion to dismiss in the Times and Tribune cases argued this directly, attempting to shift legal responsibility for specific outputs onto the people typing the questions.

That last point is where Perplexity’s defense weakens fastest. CNN’s complaint specifically alleges Perplexity continued accessing CNN content after CNN deployed technical blocks against its crawlers. A user typing a prompt does not explain a crawler defeating a block placed specifically to stop that crawler.

The “Skip the Clicks” Problem

Perplexity built its early marketing around a direct promise: skip the links, get the answer. That tagline became a liability the moment publishers needed to prove substitutive harm.

A 2025 study from content-licensing platform TollBit, cited in the Times’s complaint, found AI search engines send roughly 95 to 96% less referral traffic to news sites compared to traditional search. That is the commercial heart of every publisher’s case: not that Perplexity copied an article once, but that its entire product is engineered to replace the visit a publisher’s business model depends on.

Intellectual property attorney Steve Kramarsky observed that Perplexity quietly dropped the “skip the click” tagline after Dow Jones sued. Removing the marketing language after litigation begins does not erase what the company built its growth strategy on for the prior two years, and plaintiffs’ lawyers have entered the old tagline into the record specifically to make that point to a jury.

The Amazon Case: A Different Kind of Scraping Claim

The Amazon lawsuit stands apart from the publisher cases because it does not center on copyright at all. Amazon’s claim is about unauthorized access and concealment, not reproduction of expressive content.

Amazon’s complaint alleged Perplexity’s Comet browser disguised its AI shopping agent as an ordinary human browsing session, specifically to defeat Amazon’s bot-detection systems. Judge Chesney’s March 2026 ruling found Amazon had submitted “essentially undisputed evidence” supporting that claim.

Amazon also identified a secondary harm distinct from the scraping itself: its advertising business now has to detect and filter AI-agent traffic before billing advertisers, since advertisers pay only for verified human impressions. That is a concrete, quantifiable cost stemming directly from Perplexity’s conduct, which is part of why the injunction came relatively quickly compared to the slower-moving publisher cases.

What the Privacy Class Action Could Mean Financially

The Doe privacy lawsuit carries a different kind of exposure than the copyright cases. It seeks statutory damages of $5,000 per violation under the California Invasion of Privacy Act, not damages tied to provable actual harm.

That math does not mean the case is worth $100 billion. Statutory damages calculations get heavily contested at the class certification stage, and courts routinely narrow per-violation theories before they reach anything close to a literal multiplication of users times penalty. What matters here is that CIPA’s statutory framework removes the burden of proving individualized financial harm, which is normally the hardest part of a privacy class action to litigate.

The complaint also names Meta and Google as co-defendants, alleging both companies violated state and federal computer fraud and privacy statutes by accepting and using the transmitted data. That widens the case beyond a single AI startup into a test of how three of the largest technology companies handle third-party data flows.

Why Some Publishers Sued While Others Signed Licensing Deals

Not every publisher chose litigation. Time, USA Today’s parent company Gannett, Le Monde, and Der Spiegel all struck content-licensing partnerships with Perplexity rather than suing. CNN itself signed a separate licensing deal with Meta in December 2025, demonstrating the network embraces AI partnerships on its own terms when payment is involved.

That split reveals the real dispute is not whether AI companies may use news content at all. It is whether they pay for it. CNN’s spokesperson made the position explicit: “There is no free option.”

Perplexity’s own Publishers Program offers revenue sharing to participating outlets. CNN’s complaint specifically notes it was never invited to participate, despite its content appearing in Perplexity’s outputs regardless. That detail undercuts any argument that Perplexity treats unlicensed use as accidental or incidental to its core business model.

The Snapchat lawsuit over teen data and engagement design shares a structural theme with the Doe privacy case here: a platform’s core product depends on data flows the company never adequately disclosed to the people generating that data.

What This Means for Perplexity Users Right Now

Anyone using Perplexity, particularly its Incognito mode, should understand that the privacy protections advertised in 2025 and early 2026 are the subject of active litigation, not settled fact. The Doe complaint alleges those protections did not function as marketed.

No claims process exists yet in any of these cases. The publisher suits seek injunctions and damages payable to the corporate plaintiffs, not individual consumers. The Doe privacy case is the only one structured to eventually produce payments to ordinary users, and only if it survives motions to dismiss and reaches class certification.

What This Lawsuit Teaches Consumers

The pattern across every Perplexity lawsuit filed since October 2024 is the same: a company built a product that depends entirely on other people’s work, whether that work is journalism, retail listings, or personal data, and externalized the cost of acquiring it.

That is not a uniquely AI problem. Search engines, content aggregators, and social platforms have litigated versions of this fight for two decades. What makes the current wave different is speed and scale. Perplexity reached roughly 20 million U.S. monthly users while at least seven major publishers, one of the largest retailers in the world, and a proposed nationwide privacy class all filed federal complaints within an 18-month window.

For consumers, the lesson sits in the Doe case specifically. A feature labeled “Incognito” is a promise, not a technical guarantee, until a court or a settlement forces verification. Companies advertise privacy controls constantly. Far fewer companies get independently audited on whether those controls actually do what the marketing claims.

For publishers and content owners watching from outside this litigation, the through-line is clearer still. Perplexity already operates licensing deals with Time, Gannett, and others. The companies suing it are not arguing that AI search is inherently illegitimate. They are arguing that taking content without paying for it, while marketing a paid program you were excluded from, is the part that belongs in federal court.

Read These

• Celtic Sea Salt lawsuit
• Ed Sheeran lawsuit
• Homeland Security student visa lawsuit
• Equifax lawsuit
• Snapchat lawsuit