Media Briefing: Perplexity’s new ‘trust and transparency’ pitch does little to win over publishers

The words “trust” and “trustworthiness” came up at least a dozen times during a presentation by Perplexity’s head of publisher partnerships, Jessica Chan, at an IAB Tech Lab event last Thursday.

And yet, two and a half years after launching its publisher partnership program, Perplexity is facing an uphill battle to win publishers’ confidence. On May 28 — the same day as Chan’s IAB Tech Lab presentation — CNN filed a lawsuit accusing the AI startup of unlawfully distributing its copyrighted content. 

It was the latest in a string of copyright lawsuits publishers have brought against Perplexity. It was sued by The New York Times, Chicago Tribune, and Dow Jones last year. BBC threatened Perplexity with legal action last June, alleging the company was reproducing its content “verbatim” without permission.

Chan alluded to the tension between publishers and Perplexity during her presentation, assuring attendees that the AI search engine was committed to reestablishing its credibility.

“We know AI companies have not always earned publishers’ trust,” she said. “Trust me, I’m front and center to all of it, and I have these conversations daily with these publishers. But we know that rebuilding it will take transparency, consistency, and shared value, and that’s the direction that we are committed to.”

Perplexity’s goal, according to Chan, is for its AI answer engine to be “accurate and verifiable,” with answers built around citations and link-backs to sources.

The catch: for publishers, citation-first isn’t the same as revenue-first, and that’s ultimately what matters.

Sure, publishers are concerned about how their content appears in AI answer engines and have pushed for more and improved attribution (various licensing deals between OpenAI and publishers were centered around surfacing and citing their content in AI-generated responses) — but publishers also want money for their content. And it doesn’t seem like they’re getting much of that from Perplexity.

“Last week was a tough time for [Chan] to bring that talk in the room, given how the dynamics on all of this have changed over the past year,” said a media exec, who attended the event and agreed to speak under the condition of anonymity. “Even with that empathy, I found myself taking it all in pretty skeptically given Perplexity’s past behavior and open lawsuits.”

This is not Perplexity’s first attempt to placate publishers. It launched its publisher partnership program in July 2024 with an ad revenue share model, onboarding publishers like Blavity, Der Spiegel, Fortune, The Independent, Los Angeles Times, and Time. Perplexity then introduced a subscription revenue share model through Comet Plus last August.

Perplexity is unusual in that it hasn’t adopted more straightforward AI licensing deals or “pay per crawl” models that charge bots for accessing content, like OpenAI, Amazon, Meta and other tech companies have.

At the IAB Tech Lab event, Chan teased a new licensing program that launched in March called “premium sources,” where Perplexity signs deals with data providers to bring professional, peer-reviewed research that is “clearly cited and traceable” into Perplexity Computer, its agentic AI assistant platform. The program has already signed on companies in the health and finance categories, and will “move into areas like news a little later this year,” Chan said. 

Perplexity also gave a demo showing that its AI agents don’t just summarize editorial coverage, but can also package it for sales. Chan pointed to an example where its Perplexity Computer agents could scan a month of articles, cluster audience intent, then map that to advertiser categories to create a sponsorship-ready concept. Essentially, turning editorial patterns into sellable campaigns in minutes rather than days.

While Perplexity is trying to build a different path to publisher partnerships, it may be missing a chance to become the truly publisher-friendly AI search engine it wants to be. That may not be enough for many publishers, but at least it’s putting something on the table, the media exec said.

“[Perplexity] can’t just have deals with a few publishers and pretend that’s fair use compensation in the marketplace. That’s a PR strategy,” said the exec. “But that’s not just a Perplexity question, and I think it’s fair to say we’ve seen more movement on this from Perplexity than say, Google, who has the vast majority of AI search market share,” they added. 

Some publishing execs have told Digiday that it’s been difficult to get a hold of Chan to discuss joining Perplexity’s publisher program. Chan is a one-woman team (a source told Digiday that Perplexity had hired someone to join her, but when reached for comment, a Perplexity spokesperson simply said, “Perplexity is always hiring.”).

As one publishing exec told Digiday, “We frankly don’t feel the love.”

But a publishing exec at a company that is part of Perplexity’s publisher program found the pitch to match the current environment, where AI companies and publishers need to support each other, they said.

“Based on my experience with Perplexity, I have no reason to doubt their sincerity. At this stage, AI companies must nurture the publishing ecosystem that feeds their consumer facing AI products. Publishers need to survive for all this to work. But the change needed to get there is going to be bumpy,” they said. “Publishers and AI companies must coexist for either to survive.”

And yet, Perplexity’s AI bots have been known to ignore publishers’ protection measures in robots.txt to deter content scraping without compensation, publishing execs previously told Digiday, though Perplexity denies this.

“Perplexity is going to have to do a lot more than wave around a new program to earn publishers’ trust after their repeated flagrant violations of that trust in the past,” said another media exec, who spoke under the condition of anonymity. 

Perplexity may be attracting more publisher lawsuits than other AI search engines because its AI-generated summaries can serve as a substitute for visiting the original source, rather than driving traffic back to publishers, according to a media and tech lawyer, who requested to speak anonymously. 

Also, Perplexity doesn’t use publisher content to train its own large language model, they added. This “leaves the potentially easier claim for publishers to focus on — Perplexity built a platform that often reproduces publishers’ content in a way that acts as a substitute for visiting those sites, as opposed to a search tool that drives traffic to those sites,” they said.

Chan insists “strong AI and strong publishers” should reinforce each other, but after Perplexity’s recent breaches, media companies are still waiting to see that sentiment show up in contracts, and as reliable licensing revenue, not just keynotes. Until that happens, publishers say the trust gap remains wide open.

What we’ve heard

“When we made that move from that block list to that allow list, some of the numbers were just really staggering in terms of how many bots there were that were out there, and we’re still not catching all of them. We went from blocking roughly 2,100 user agents… to over 30,000… It gives you a sense of just how big of a scale this challenge really is.”

— Lindsay Van Kirk, People Inc’s svp of innovation, speaking onstage at an IAB Tech Lab event last Thursday.

Numbers to know

$20 million: The amount The New York Times has spent on legal fees in its lawsuit against OpenAI over the past two and a half years.

$18 billion: The amount People Inc. is offering to buy casino giant MGM Resorts.

Over $400,000: The amount Condé Nast is paying three journalists to settle a dispute over their firing last fall.

Tens of millions: The revenue Axios’s events business now generates.

What we’ve covered

The case for and against publishers buying paid traffic 

• Publishers are quietly leaning harder on paid traffic to offset referral traffic declines, treating it as a lifeline for audience growth in a world where search referrals may eventually “go to zero.”
• For many audience development teams, the question is no longer whether to buy traffic, but how far they can push it without tipping into the kind of arbitrage that buyers and made-for-advertising (MFA) blocklists now punish.

Read more here.

How publishers are modeling – and mitigating – a future with less Google search traffic

• Publishing execs aren’t convinced if — or when — Google Zero will happen, but they’re preparing for the worst anyway: a future where Google search stops driving traffic to their sites, and becomes a destination for information.
• Publishers including Condé Nast and Time are strategizing for Google referral traffic tanking. Digiday spoke to six execs at large publishers about how they’re modeling for less Google traffic.

Read more here.

Publishers cut ‘six-figure’ deals via Snowflake’s AI licensing platform

• Publishers are quietly cutting six-figure AI licensing deals on Snowflake, as the data giant positions itself as matchmaker-in-chief between locked-down news content and enterprises keen to plug reliable publisher content into their own internal AI tools via retrieval-augmented generation (RAG).
• Its a way to monetize the RAG pipe, by letting enterprises query their paywalled or proprietary content inside Snowflake’s AI environment without exposing raw feeds, losing control, or getting scraped for model training. The Washington Post, AP, People Inc., and USA Today Network are among the 17 publishers that have signed on to use it.

Read more here.

Vox Media CRO Geoff Schiller joins Screenvision as CEO

• Geoff Schiller, former CRO at Vox Media, is set to become CEO of cinema ad network Screenvision. The move marks Schiller’s first departure from the digital publishing industry in 25 years.
• Schiller told Digiday the move wasn’t because he’s soured on digital publishing, but because he was drawn to leading Screenvision at an interesting time for the theater industry, given Gen Z’s interest in moviegoing and a strong pipeline of films in the next few years.

Read more here.

The healthcare creator is finally diagnosing how they best fit into the creator economy

• As more people look for ways to be healthier or seek information about medical ailments, creators in this space have stepped up. They’re becoming increasingly valuable to brands looking for credibility and social media influence.
• However, these creators face stricter regulatory and ethical standards, making brand partnerships more expensive and complex.

Read more here.

What we’re reading

U.K. publishers can now choose not to appear in Google’s AI Overviews

The UK watchdog org the Competition and Markets Authority (CMA) has announced publishers can now choose not to appear in Google’s AI Overviews, giving more control to publishers about their content appearing in AI products, the BBC reported. Tests will happen in the UK first before rolling out globally. 

YouTube exec teases developing product to integrate publisher paywalls

YouTube is working on integrating its platform with publisher paywalls, YouTube’s vice president for Europe, Pedro Pina, said at the recent WAN-IFRA World News Media Congress event in Marseille. It would allow publishers to put videos under a paywall up on YouTube, A Media Operator reported.

Glamour will soon be not much more than a site of shopping links

Glamour is pivoting from a more traditional, lifestyle magazine publisher to a site largely focused on shopping posts aimed at driving affiliate commerce revenue, The New York Times reported. 

Barry Diller sees casino business as hedge for People Inc’s publishing arm

People Inc chairman Barry Diller, now 84 years old, wants to take over the rest of MGM Resorts, as a way to insulate the company’s publishing business in a rapidly evolving, AI-driven media era, The Wall Street Journal reported.

AP and OpenAI sign deal on election data

The Associated Press and OpenAI have struck a deal that will license the wire service’s elections data through the 2028 U.S. presidential elections, Variety reported. The two companies first signed an AI licensing deal in 2023.