Cable News Network yesterday filed a landmark copyright and trademark lawsuit against Perplexity AI in the United States District Court for the Southern District of New York, alleging the AI search company unlawfully copied and redistributed more than 17,000 CNN stories, videos, images, and other works to power its commercial products - without authorization or payment.

gov.uscourts.nysd.664916.1.0_1
gov.uscourts.nysd.664916.1.0_1.pdf
15 MB
download-circle

The case, Cable News Network Inc v. Perplexity AI, Inc. (1:26-cv-04427), was filed on May 28, 2026, by law firm Rothwell, Figg, Ernst & Manbeck, P.C., with attorney Steven Lieberman signing the complaint. A filing fee of $405.00 was paid on submission, and the complaint is accompanied by two exhibits: Exhibit A listing the CNN works at issue and Exhibit B setting out CNN's registered trademarks.

gov.uscourts.nysd.664916.1.1
gov.uscourts.nysd.664916.1.1.pdf
7 MB
download-circle
gov.uscourts.nysd.664916.1.2
gov.uscourts.nysd.664916.1.2.pdf
7 MB
download-circle

What CNN alleges

According to the complaint, Perplexity's conduct violates CNN's exclusive rights under the Copyright Act at two distinct stages. At the input stage, Perplexity allegedly crawls, scrapes, copies, and distributes CNN content from CNN's digital platforms and third-party platforms using software programs including "PerplexityBot" and "Perplexity-User," to build what Perplexity describes as an "AI-First" search index. At the output stage, Perplexity's products then generate responses to user prompts that are, according to the filing, "identical or substantially similar to CNN's content."

The complaint states that Perplexity has "unlawfully copied over 17,000 CNN stories, videos, images, and other works to power its products and tools." CNN owns registered copyrights in those works, primarily consolidated under registration TX 9-574-794, which the complaint cites repeatedly in examples of verbatim reproduction.

CNN is a Delaware corporation with its principal place of business in Atlanta, Georgia, and is a wholly owned subsidiary of Warner Bros. Discovery, Inc. Perplexity is a Delaware corporation headquartered at 115 Sansome Street, Suite 900, San Francisco, California 94104. According to the complaint, Perplexity expanded its Manhattan office footprint to approximately 22,000 square feet in February 2026, and as of May 2026 listed 23 open positions based in New York.

How Perplexity's retrieval system works

The complaint goes into considerable technical detail on the mechanics of retrieval-augmented generation, commonly called RAG, which lies at the center of CNN's allegations. According to the filing, Perplexity has built a search index comprising "hundreds of billions of webpages" that it crawled from the internet. Perplexity's own public statements describe a "crawler and indexing fleet" comprising "tens of thousands of CPUs and hundreds of terabytes of RAM," with an "exabyte-scale index and crawling apparatus."

RAG works in four steps, according to the complaint: the system receives a user prompt; retrieves relevant content related to that prompt; combines the original prompt with the retrieved documents; and feeds that combined data to a large language model, which generates a human-like response. CNN argues that this process requires Perplexity to make copies of source content at two legally distinct points - once when ingesting it into the index, and again when serving it as output.

Perplexity's chief executive Aravind Srinivas is quoted in the complaint describing the crawling cycle as "pretty frequent, like at least every few hours." The company has also said publicly that its "content is sourced from the web in real-time as you ask your questions," and that "all responses are supported by citations from reputable news organizations, academic publications, and established content sources."

The complaint also describes Perplexity's proprietary "Sonar" family of large language models, which were built on Meta's open-source Llama 3.3 70B model and fine-tuned specifically to support "real world use cases" including "browsing news, sports, health and finance content." As of February 11, 2025, Perplexity made Sonar available as the default model for its API.

The robots.txt evasion allegations

A substantial section of the complaint concerns Perplexity's alleged circumvention of technical protections. CNN's complaint states that Perplexity openly acknowledges that its "Perplexity-User" crawler "generally ignores robots.txt rules." The complaint further cites a Cloudflare investigation published August 4, 2025, in which four independent researchers documented that Perplexity employed "stealth, undeclared crawlers" that impersonated Google Chrome on macOS and "utilized multiple IPs not listed in Perplexity's official IP range" to access content even when websites had explicitly blocked it.

Those findings align with earlier documentation of Perplexity's crawling practices that PPC Land covered when Cloudflare first documented stealth crawler activity in 2024, and with the October 2025 Reddit lawsuit that named Perplexity as a defendant for similar conduct.

CNN's complaint states that CNN had implemented robots.txt blocks against PerplexityBot, yet Perplexity continued accessing CNN content. The complaint also states that a Conde Nast engineering analysis of its system logs found that Perplexity's undisclosed IP address had "likely accessed the company's content thousands of times without permission."

The failed licensing negotiations

One of the more commercially revealing sections of the complaint concerns a licensing attempt that ultimately collapsed. According to the filing, CNN and Perplexity negotiated a term sheet effective October 1, 2025, for a "CNN - Perplexity Comet Plus Partnership." On October 2, 2025, Perplexity publicly announced the agreement, with CEO Srinivas telling Business Insider: "We have always been clear that a product like ours required high-quality sources to exist on the web."

The parties were unable to finalize the contract within the agreed 60-day window, primarily because of disagreements over limits on Perplexity's use of CNN content in answers to users. Both parties agreed to terminate the term sheet effective November 24, 2025.

On December 10, 2025, CNN wrote to Perplexity's Head of Legal and Operations Nathan Barksdale demanding that Perplexity cease use of CNN's trademarks and content. According to the complaint, Perplexity never responded to that letter and continued using CNN content. The lawsuit followed approximately five months later.

Verbatim reproduction examples in the complaint

The complaint presents specific side-by-side comparisons showing Perplexity output alongside original CNN text, with allegedly copied passages marked in red. In one example from May 2026, when a user asked Perplexity Pro to "deeply analyze an article from a single reputable source" about Marco Rubio's job titles, the system produced text that the complaint describes as closely tracking a CNN article by Aleena Fayaz published February 5, 2026. That article is registered as part of Copyright Registration TX 9-574-794.

A second example involves Perplexity's Comet browser assistant, which the complaint alleges produced verbatim text from a paywalled CNN article by Stephen Collinson published February 9, 2026, even though CNN's page displayed a paywall and required a subscription to access. The complaint presents two screenshots showing the CNN subscription prompt on one side of the screen and the Comet assistant reproducing article text on the other.

The complaint further shows output from Perplexity's Search API, priced at $5 per 1,000 requests, reproducing substantial verbatim portions of multiple CNN articles in response to title-based prompts. In the case of the Agent API, the complaint alleges a user was able to obtain the entire verbatim text of a CNN article by prompting Perplexity to perform a "sentiment analysis," with an instruction to "try other approaches for accessing it" if direct access failed.

According to the complaint, when using the free tier, Perplexity responds to requests for full article text by saying "I can't provide the full text of that article, but can offer a detailed summary or quote a short excerpt." However, the paid API tiers do not appear to apply the same restriction.

The trademark claims

Beyond copyright, CNN brings claims under the Lanham Act alleging trademark infringement, false designation of origin, and dilution. The complaint states that Perplexity's chatbot, when asked about "Comet Plus," generated a response falsely telling users they could obtain CNN premium content through Perplexity's subscription service - a partnership that had not been finalized and no longer existed as of the lawsuit date.

Exhibit B to the complaint contains 11 trademark registration certificates for CNN marks spanning International Classes 9, 35, 38, 41, 42, and 45. The oldest registration cited is Reg. No. 1,597,839, registered May 22, 1990, for cable and television broadcasting services. The most recent is Reg. No. 5,860,237, registered September 17, 2019, covering a downloadable mobile application. Other marks included are CNN HEROES, CNN FILMS, CNN SANS, CNN NEWSOURCE, CNN COLLECTION, WATCH CNN, and CITIZEN BY CNN.

The complaint also presents an example where Perplexity attributed garbled, nonsensical statements to CNN's coverage of Arsenal football manager Mikel Arteta - including a fabricated quote reading "I burn every of blood this football to make it better" - while displaying the CNN trademark alongside the hallucinated content.

Scale and financial context

According to the complaint, Perplexity was valued at $20 billion in its most recent funding round and has raised nearly $1.5 billion in total. As of November 2025, the platform had approximately 22 million active users. In May 2025, Perplexity's CEO stated the platform had handled 780 million search queries that month. On January 29, 2026, it was reported that Perplexity signed a $750 million cloud services agreement with Microsoft.

One study cited in the complaint found that "AI bots on average are sending 95.7% less referral traffic than traditional Google search." CNN's complaint frames the traffic diversion as an existential threat to news economics: with less revenue, news organizations will have fewer journalists able to dedicate time and resources to important in-depth reporting, which creates a risk that those stories will go untold.

According to the complaint, in 2024 Perplexity spent $48 million on cloud services, $19 million on talent, and $8 million paying Anthropic and OpenAI to use their models, while paying CNN nothing.

CNN maintains 36 bureaus around the world and employs approximately 3,500 people, the complaint states. In 2025, CNN Digital Platforms reached an average of 94 million monthly unique users in the United States, making it the first-ranked US news outlet in digital reach for the tenth consecutive year. The network had 67 million average monthly mobile visitors.

Reaction and Perplexity's defense

In a statement cited by CNN's own reporting, a CNN spokesperson said: "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 Perplexity exploits. 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."

Perplexity's chief communications officer Jesse Dwyer responded with a four-word statement: "You can't copyright facts." According to CNET's reporting on the lawsuit, Michael Goodyear, an associate professor at New York Law School, noted that while facts themselves are not protected by copyright, the way CNN presents facts could be. "Even short news articles would typically qualify for copyright protection under the low bar of required originality," Goodyear said, adding that the key question is "whether the thousands of cases of infringement CNN describes are copying whole paragraphs verbatim, or whether they are paraphrasing or merely copying unprotectable facts."

The lawsuit is also related to two earlier cases already before the same court: the New York Times case (No. 1:25-cv-10106) and the Chicago Tribune case (No. 1:25-cv-10094), both filed in December 2025. CNN filed statements of relatedness in its docket entries for both.

Prior litigation pattern

CNN's lawsuit is not the first Perplexity has faced over content practices. PPC Land covered the Encyclopaedia Britannica and Merriam-Webster case filed September 10, 2025, which made structurally similar allegations including RAG copying and trademark violations. The Dow Jones & Company lawsuit (No. 1:24-cv-07984) was filed in late 2024. The October 2025 Reddit lawsuit focused on circumvention of technological controls rather than copyright per se. Amazon sued Perplexity in November 2025 over the Comet browser's unauthorized access to Amazon's systems, a case that attracted a coalition of major publishers filing an amicus brief in the Ninth Circuit in April 2026.

In August 2025, the Dow Jones court denied Perplexity's motion to dismiss in full, finding that the company's "extensive contacts with this state - such as hiring key employees in New York, leasing office space in New York, and targeting New York users with advertisements and New York-specific webpages" supported jurisdiction. That ruling provides a framework that CNN's attorneys appear to have relied upon in crafting their own jurisdictional arguments.

Why this matters for the marketing and advertising industry

For the marketing community, the CNN lawsuit adds legal weight to a structural question that sits at the heart of AI search economics: whether AI platforms can build commercially viable products by summarizing the output of professional news organizations without paying for it.

The complaint's financial disclosures are unusually detailed. The assertion that Perplexity spent zero dollars on CNN content while generating $750 million Microsoft deals and a $20 billion valuation frames the dispute in terms that advertisers and media buyers understand: the value chain from journalism investment to AI answer generation is currently extractive in one direction.

The complaint also raises a specific concern for paid search and programmatic buyers. CNN's digital platforms generate revenue from advertising, and Perplexity's CEO has publicly described the platform's goal as eliminating "extra steps and clicks" - which translates directly to reduced page visits, reduced ad impressions, and reduced publisher revenue. According to the Open Markets Institute report cited in CNET's coverage, the rate of AI crawlers bypassing paywalls and blocks nearly quadrupled over six months, rising from 3.3% to 12.9%.

The AI crawling-to-referral imbalance that PPC Land documented in September 2025 showed Perplexity increasing its crawl intensity 256.7% relative to referrals over the first half of 2025, climbing from 54 crawls per referral in January to 195 by July. That data predates the CNN lawsuit and provides independent corroboration of the traffic diversion pattern that CNN alleges.

For publishers that rely on a combination of licensing deals and traffic-driven advertising, the lawsuit signals an increasingly binary market: either negotiate a formal content agreement or litigate. CNN's own statement made that explicit: "There is no free option."

Timeline

Summary

Who: Cable News Network, Inc., a wholly owned subsidiary of Warner Bros. Discovery, Inc., filed the lawsuit as plaintiff. Perplexity AI, Inc., a Delaware corporation headquartered in San Francisco and valued at $20 billion, is the defendant. Attorney Steven Lieberman of Rothwell, Figg, Ernst & Manbeck, P.C. filed on CNN's behalf.

What: A federal complaint alleging copyright infringement under 17 U.S.C. § 106(1) and § 106(2), contributory and vicarious copyright infringement, false designation of origin, trademark dilution under 15 U.S.C. § 1125, and trademark infringement under 15 U.S.C. § 1114. CNN alleges Perplexity copied more than 17,000 CNN works for use in its RAG-based AI products - including its consumer answer engine, the Comet browser assistant, the Search API, and the Agent API - and that Perplexity's outputs falsely attributed content and affiliation to CNN. CNN seeks injunctive relief, statutory damages, actual damages, treble damages on the trademark counts, restitution of profits, and attorneys' fees.

When: The complaint was filed on May 28, 2026. The alleged infringement dates back to Perplexity's launch in 2022. Direct communications between CNN and Perplexity regarding the unauthorized use began no later than December 10, 2025.

Where: The United States District Court for the Southern District of New York, where Perplexity is already a defendant in at least five other lawsuits. Perplexity rents 22,000 square feet of office space in Manhattan. CNN operates studios at 30 Hudson Yards in New York with more than 800 staff.

Why: CNN argues that Perplexity built a commercial AI search product with a $20 billion valuation by systematically copying professional journalism without payment, while its products are designed to eliminate the user's need to visit original news sources - directly reducing CNN's subscription, advertising, licensing, and affiliate revenue. The broader context is an industry-wide conflict over whether AI answer engines can extract value from journalism investment without contributing to the economic model that makes that journalism possible.

Share this article
The link has been copied!