Perplexity faces new legal pressure even as it attempts to win over publishers and advertisers

The complaint also alleges Perplexity answers user queries in ways that plagiarize paragraphs or whole articles without linking to the original source. Another claim alleges Perplexity’s answers violate News Corp trademarks through “hallucinations” that falsely attribute information to The Wall Street Journal and The New York Post. The issues result in what the complaint describes as a “massive amount of illegal copying” that diverts away customers and revenue. 

The lawsuit comes as The New York Times is also applying legal pressure on Perplexity. Earlier this month, the Times sent Perplexity a cease-and-desist letter over similar issues related to scraping paywalled content and including content in answers without permission. 

“Perplexity’s use of these marks in relation to hallucinations is likely to cause dilution by blurring and/or tarnishing Plaintiffs’ famous marks,” according to the lawsuit. “In addition, these false designations of origin confuse and deceive Perplexity users into believing that the hallucinations are associated with, sponsored by, or approved by Plaintiffs, when they are not. This, in turn, causes significant harm to Plaintiffs.”

The legal battles come as the news industry continues its bifurcated approach to potential opportunities and threats posed by AI search platforms. Numerous publishers have announced new deals with Perplexity, OpenAI and Google that tout new revenue and AI product development. Other publishers are taking the road that leads to court. Since last year, news publishers and authors have filed other lawsuits against OpenAI, Microsoft and Anthropic. 

Among the deal-makers is News Corp, which in May signed a multi-year deal with OpenAI reportedly valued at $250 million. Other deals this month include a $10 million deal announced between OpenAI, Microsoft and several publishers including The Minnesota Star Tribune, Philadelphia Inquirer and other major metropolitan newspapers. OpenAI also announced a separate deal this month with Hearst that includes more than 20 magazines and 40 newspapers.

Perplexity has worked to make inroads with publishers and advertisers. In July, it announced an ad revenue-sharing program for publishers — including Time, Fortune, Der Spiegel and WordPress parent company Automattic — in exchange for using quality news content. Its stated goal is to sign dozens before the end of the year. A month later, Perplexity began circulating a pitch deck to advertisers with the plan of introducing ads to the platform and across third-party pages. (One page in the pitch deck reviewed by Digiday displayed a Wall Street Journal article with examples of ads that might be bought through Perplexity.)

Perplexity did not immediately respond to Digiday when asked for comment about the lawsuit. However, a News Corp spokesperson provided a statement to Digiday from News Corp. CEO Robert Thomson, which described Perplexity’s actions as “content kleptocracy.” He said the AI startup “perpetrates an abuse of intellectual property that harms journalists, writers, publishers and News Corp.”

“The perplexing Perplexity has willfully copied copious amounts of copyrighted material without compensation, and shamelessly presents repurposed material as a direct substitute for the original source,” according to the statement. “Perplexity proudly states that users can ‘skip the links’ – apparently, Perplexity wants to skip the check.”

Speaking at a Wall Street Journal event on Wednesday, Perplexity CEO Aravind Srinivas said he was “very surprised about the lawsuit because we actually wanted a conversation.” According to the Journal’s report from the event, Srinivas said the advertising program will launch later this month and likened Perplexity’s revenue-sharing plan to Spotify’s pay model for artists.

Will legal battles keep advertisers from dipping their toes into potentially hot water? It’s hard to say. AI lawsuits haven’t yet hindered startups from attracting new customers or investors, noted Edward Klaris, a managing partner at Klaris Law and an adjunct professor at Columbia Law School. 

“The average person, whether it’s an advertiser or a consumer or an investor, is not really feeling like there’s a massive risk here,” said Klaris, who was an attorney for ABC News and The New Yorker before working on media deals for nearly a decade at Conde Nast. “So my instinct is that as long as it’s useful to the advertiser to be there and they’re going to get the kind of eyeballs they want, they’re going to advertise…That’s just the economic environment we’re in.”

Publisher lawsuits and deals related to AI can comfortably coexist, Klaris said. He also noted AI companies won’t necessarily be quick to settle even if publishers use legal complaints as a means to licensing deals. That’s a potentially pricey legal bet that not all publishers would be willing to make: “You can’t sue for some hypothetical just because there’s stuff in the training data.”

Beyond ad revenue, others note there could be other ways to partner with AI companies on business models. Mark Zamuner, CEO of Juice Media, mentioned some sort of affiliate model where a publisher gets credit and some sort of revenue when an AI platform uses its content in answers. That might be helpful for areas like e-commerce when it comes to product information and recommendations. The ethical collection, distribution and attribution of content is key. 

“If you’re not attributing to sourcing, that becomes problematic,” said Zamuner. “Because you can do one of two things. You could disintermediate while pulling from an actual source that should get credit, or you could just be in a disinformation case like you’re just taking a marketing brand’s marketing output unbeknownst to the end user and monetizing it.”

It’s still hard to know the volume of outputs that might violate various trademark laws. Earlier this month, the AI detection startup Copyleaks conducted its own tests on paywalled articles from Forbes and The Information. According to the analysis, Perplexity paraphrased 48% of sample articles and in one instance plagiarized 7%. However, Copyleaks noted other AI platforms refused to summarize paywalled content.

“Many publishers currently only see the downside of having their content scraped by Perplexity and other AI-first search engines,” said Jim Yu, founder and executive chairman at BrightEdge, an enterprise SEO platform. “The upside in the form of referrals is extremely small due to these new platforms’ limited market share. However, it is about the long game. These engines are growing at an impressive rate and will likely soon have a sizable impact on the search market.”