The pros and cons of publishers’ AI licensing deals

As a Digiday+ member, you were able to access this article early through the Digiday+ Story Preview email. See other exclusives or manage your account.This article was provided as an exclusive preview for Digiday+ members, who were able to access it early. Check out the other features included with Digiday+ to help you stay ahead

Are you even a large digital publisher if you haven’t signed a deal with OpenAI?

Last week, the generative AI tech company announced content licensing and tech development agreements with Vox Media and The Atlantic. This comes after deals were recently announced between OpenAI and IAC’s Dotdash Meredith, as well as The Wall Street Journal’s parent company News Corp — following deals with The Financial Times, Axel Springer and The Associated Press.

There are a few large digital publishers that haven’t yet come aboard — namely Condé Nast, Hearst and Bustle Digital Group. The New York Times is taking a different approach entirely, suing OpenAI and Microsoft in Decemberas is a group of newspapers owned by Alden Global Capital.

The recent wave of deals with OpenAI means publishers are getting compensated for their content being used to feed and train large language models (LLMs). This has been a major sticking point for publishers as these large tech companies have scraped their content without permission or payment.

But the agreements are also getting a fair amount of criticism. Compensation for use of content is good for a publisher, but are they getting paid enough? Are publishers signing away their most valuable assets, willy-nilly?

“The recent agreements between AI developers and publishers represent a positive step toward recognizing the critical role publishers play in AI training,” said Francesco Marconi, a computational journalist and co-founder of real-time information company Applied XL, which has a partnership with Bloomberg Media. “However, it’s important to consider the long-term viability of these partnerships. Once an AI company licenses an archive and completes initial model training, there is little incentive for the company to renew data access agreements.”

Digiday has compiled a non-exhaustive list of the pros and cons of media companies signing deals with AI tech companies like OpenAI. Here’s a look at some of them:


Publishers can monetize their archives and content by AI companies paying them to use it to train their LLMs

The most obvious benefit to publishers is getting paid by tech companies for the use of their content, rather than using it for free.

“We haven’t been shy about our belief that publishers should be compensated for use of their content and that attribution and transparent sourcing in generative AI products is critical to maintaining a healthy Internet ecosystem. This partnership addresses both fronts,” IAC CEO Joey Levin said during the company’s Q1 earnings call on May 8.

While publishing execs have not said how much they’re getting paid by OpenAI, The Information reported the tech company has been offering some publishers $1 million to $5 million a year to license their content, while News Corp’s five-year deal with OpenAI is reportedly valued at as much as $250 million in cash and OpenAI credits. (In comparison, payouts from Facebook to news publishers to distribute their content on its News tab ranged from the single to double digit millions — before those efforts were shut down entirely in 2022.) 

“In the race to train the largest foundation models, well-structured data which news publishers can provide is very useful. Likewise, quality data that is up to date is also needed if models are supposed to provide timely answers,” said Felix Simon, a research fellow in AI and news at Oxford University who studies the implications of AI for journalism.

AI technologies can enhance analytics, content optimization and ad targeting capabilities

These deals can also benefit publishers by giving them close access to powerful AI technology, which can help them develop products for their own employees, audience and advertisers. This can be especially helpful at a time when publishers are testing generative AI products built internally.

For example, OpenAI tools power Vox Media’s affiliate commerce product The Strategist Gift Scout, a search tool that matches users with curated gifts. Going forward, Vox Media’s deal with the AI tech company will help improve the creative optimization and audience segment targeting for the publisher’s first-party data platform Forte, according to the company.

The Atlantic’s statement announcing the OpenAI deal noted that access to the tech company’s tools will give the publisher the ability to collaborate on products that can benefit its journalists.

Excerpts featured in generative AI chatbot responses could increase traffic

Publishers will have more of a say in how their content is used and surfaced by generative AI chatbot products like OpenAI’s ChatGPT.

The Atlantic’s articles will be discoverable within ChatGPT and the publisher “will help to shape how news is surfaced and presented in future real-time discovery products,” according to the company’s statement. Queries that surface The Atlantic’s content will include attribution and a link to read the full article on the publisher’s site, the company added.

The hope is that this kind of attribution can help drive traffic back to publishers’ sites when their content is referenced in answers to prompts in ChatGPT. With the rise of generative AI, publishers have expressed concern that chatbots like ChatGPT (and Google Search’s new generative AI feature) could cut their referral and search traffic, an already challenged area for many.


AI companies are underpaying and taking advantage of publishers’ lack of leverage 

But how much is a publisher’s content really worth? Is a couple million dollars enough? 

While it remains unclear what exactly OpenAI is asking for in its offers to publishers, a couple million dollars seems like a lowball amount.

Important questions, according to Simon, are if these deals are exclusive, what publishers get compensated for and how much, if the terms can be renegotiated and if they cover products and services that potentially compete with publishers’ own offerings. 

“More generally, questions arise around which and whose date is seen as most valuable and who gets access to such deals in the first place,” Simon said.

In IAC’s recent earnings call, Levin noted that the OpenAI deal is “not exclusive,” suggesting the company has the opportunity to engage in talks with other LLMs. “Hopefully that agreement is just the beginning of other opportunities for us in that AI ecosystem,” he said.

In an op-ed (ironically published on The Atlantic’s site just a week before the announcement of its deal with OpenAI), The Information’s founder Jessica Lessin addressed some of these issues. In her view, OpenAI’s deals with publishers is its way to avoid lawsuits like the one brought by The New York Times. 

“These deals amount to settling without litigation. The publishers willing to roll over this way aren’t just failing to defend their own intellectual property — they are also trading their own hard-earned credibility for a little cash from the companies that are simultaneously undervaluing them and building products quite clearly intended to replace them,” she wrote.

And that’s not to mention the potential loss of control over how their content is used and repurposed by AI companies and their products. What happens if a publisher and OpenAI can’t agree on renewal terms? If the LLMs have had access to a publishers’ decades worth of content archives already and have been trained on that data, perhaps the only leverage a publisher will have in that case is new content — valuable, yes, but how valuable? (This was reportedly one of the main issues that broke down negotiations between OpenAI and The New York Times.)

Publishers are helping to develop AI tools that could eventually replace them

It’s one thing to make it more accessible for ChatGPT users to find publishers’ content. But will it really lead to more traffic and benefit publishers? It remains to be seen whether people actually click those links (look at what happened when Google started answering queries on the search page).

Yes, publishers are getting new payouts — but feeding content to another platform doesn’t often bode well for a publisher’s ultimate goal to build, maintain and grow a relationship with their audience. It’s why sites like The Verge are doing more now to focus on that relationship, such as building a new homepage and setting their sights on lofty distribution plans.

In her op-ed, Lessin said it was “too early” for publishers to be licensing their content for “relative pennies” and called for patience.

“Technology companies aren’t in the business of news. And they shouldn’t be. Publishers have to stop looking to them to rescue the news business. We must start saving ourselves,” she wrote.

The way these deals are structured lack transparency

The lack of transparency around how these deals are structured can leave executives at other publishers in the dark, with some execs wondering if they will be left behind in the race to get money and build helpful tools to move their companies forward in the AI era.

“While major news organizations may have leverage in negotiating compensation, smaller players risk being overshadowed, leading to disparities within the industry,” Marconi said.

So far, there’s been a lack of alternatives to these deals, too — aside from publishers announcing partnerships with AI tech vendors like LedeAI and Jasper.

“We don’t know really how these deals are structured apart and for now this remains a closely guarded secret. We have seen some reported figures how much they are worth… but these are difficult to independently verify. Apart from a few general statements about what is included, we do not really know that these deals actually look like — they are very much a black box. This makes it difficult to provide a good assessment of what these deals actually mean and who is more likely to benefit from them,” Simon said.

More in Media

AI Briefing: Senators propose new regulations for privacy, transparency and copyright protections

A new bill called the COPIED Act aims to pass new transparency standards to protect IP and guard against AI-generated misinformation.

Media Briefing: Publishers reflect on ad revenue midway through 2024 

Some publishers say ad revenue is pacing 15% up year over year while others are still managing their expectations for how 2024 will shake out.

Teads is exploring sale options as M&A in ad tech heats up

Sources state the Altice-owned stalwart of outstream video has recently held talks with private equity and strategic players.