Is OpenAI funding journalism?
Media companies are making deals with OpenAI while Google bullies them into compliance.
Image provided with gobs of irony from Dall-E.
Google is starting to look more and more like Darth Vader. Stick with JMM here.
Originally, Google helped make sense of the internet, bringing order to the chaos of the open web. And it was great for everyone. We got good content. Google got some revenue from ads and sponsored links. And publishers got traffic, allowing them to sell their ads as well.
But somewhere along the line, Google went from being the middleman, connecting users to content, to being a dictator. It made publishers remake the web to meet its standards, with the caveat that Google could change those standards at any moment. It hoovered up every ad dollar it could find—over $237 billion last year alone. And it neglected its core product, letting the search engine results page (SERP) get clogged up with SEO junk and sponsored links designed to maximize revenue.
It's left consumers and publishers feeling like Lando Calrissian at the end of The Empire Strikes Back, shouting “This deal is getting worse all the time.”
And Darth Google? Well, like Vader, it’s altering the deal even more. With AI overviews, Google provides quick summaries of publisher content, greatly increasing the possibility of zero-click searches—i.e. those that don’t require you to leave Google’s walled garden to get your answer. Publishers are, of course, pissed. They’re losing traffic. Their content is being swallowed without permission.
But they also can’t quit Google. According to online marketing powerhouse Semrush, Google accounts for over 91 percent of the global search engine market. The only way to keep your content from being scrapped by Google and repurposed for its AI overviews—as well as fed into Gemini, its large language model—is to opt out of being indexed, meaning your content would never appear on any search results ever again. So, publishers are stuck with a Faustian bargain, giving up their content for diminishing returns.
OpenAI has taken a different approach. While it’s arguable that OpenAI—and every other with an LLM trained on content from the web—violated gobs of copyright laws, Sam Altman et al aren’t trying to bully anyone into handing over their content. Instead, OpenAI has been quietly inking deals with publishers for access to their latest work, ensuring that ChatGPT is continually trained on the latest high-quality content available. According to a leaked slide deck, those who sign will be a part of what OpenAI is calling the Preferred Publishers Program (PPP). As of last week, 70 publishers have signed up. The Associated Press was the first. Vox and The Atlantic are the most recent.
These deals, at least on their surface, seem to benefit both OpenAI and the media. First, publishers get a guaranteed payment—essentially a licensing fee—as well as a variable fee based on user interactions with the publisher’s content. And those payments aren’t small. A recent deal with News Corp., home of the Wall Street Journal and the New York Post, has OpenAI paying the company $250 million over five years.
That’s ridiculously better than what Google is offering, which is nothing. In fact, Google has recently penalized California media outlets, removing links to their content from search results, over the mere possibility the state could pass the Journalism Preservation Act, a bill that would require tech giants to pay a pittance to fund journalism in the Golden State. Those links are back for now, but the message is clear: Darth Google does not share.
In contrast, a press release from The Atlantic about its deal with OpenAI positions it as a real partnership. OpenAI’s Chat-GPT will not only have access to The Atlantic’s latest content but will also source it when it’s used to answer user queries, as well as provide links to complete articles. Its product engineers will also work with The Atlantic’s product team to develop new micro-sites and tools that are targeted at niche audiences, a play to increase both audience loyalty and publisher revenue. And The Atlantic will have a say in how real-time AI assistants like Chat GPT-4o use news.
It seems to be OpenAI’s standard offer. The News Corp. deal came with the same promises. So too did the one inked with the AI skeptic Dotdash Meredith, home of People, Better Homes & Gardens, and Travel & Leisure, OpenAI agreeing to help develop D/Cipher, DDM’s intent-based ad-targeting product.
"We have not been shy about the fact that AI platforms should pay publishers for their content and that content must be appropriately attributed," said Neil Vogel, CEO of Dotdash Meredith, in a press release. "This deal is a testament to the great work OpenAI is doing on both fronts to partner with creators and publishers and ensure a healthy Internet for the future."
Of course, there are skeptics about these partnerships. Some like Jessica Lessin, founder of the tech blog The Information, think publishers are giving in rather than fighting, and in doing so they’ll live to regret these deals like they regret their relationship with Google. And there are still those who are suing OpenAI over copyright infringement including the New York Times, the Chicago Tribune, the St. Paul Pioneer Press, and others. Those lawsuits will take years to play out. And while JMM is rooting for them, even if those companies win, Chat-GPT and other LLMs have already swallowed the content. It’s impossible to somehow pull it out of all that digital soup.
At least by inking these deals, media companies are getting some value out of their content. Two decades ago, they handed it over willingly to Google and Facebook, who then made billions selling ads against it without having to produce any content themselves. The PPP seems to give publishers a seat at the table for the coming AI change—emphasis on “seems.” If we’ve learned anything from Google and Meta, it’s that tech companies can change the deal to suit them at a moment’s notice, leaving the rest of us to mumble another Star Wars quote: “I’ve got a bad feeling about this.”
Summer Listening
JMM is big on podcasts. There’s always something blowing into those earholes. Recently, it’s been two offerings from Brian Morrissey, the former president and EIC of Digiday. He’s at the center of both The Rebooting, where he discusses the current state of media with execs and insiders, and People Vs. Algorithms, which includes former Global President of Hearst Magazine Troy Young and former Chief Design Officer of AirBnB Alex Schleifer. Both podcasts break down what’s currently going on in media, how brands plan to weather it all, and the options they have to profit from all this change. It’s like having a little window into media’s C-suites.
Internships, Fellowships, and More
• The New York Times is offering a Local Investigations Fellowship. It’s a year-long program that will teach both investigative techniques and local voice integration. Get more info here.
• Politico is already shopping for fall interns. Positions include editorial and photo positions in Washington, D.C., as well as editorial positions in New York and California. You’ll work 40 hours a week on Politico’s magazines, fact-checking and creating content, plus get $23 an hour. The positions run from September to December. You can apply here. The deadline for applications is June 14.
• The 30 newsletter just posted a massive list of jobs available, including a freelancer music-writer gig for Nylon that JMM has his eye on.
Got Some Fresh Dirt?
Do you have some essential info or did you just write an article for Woman’s Day that finally helps JMM to figure out if he’s a “Dark Winter” or “Dark Fall” like Kate Franke (Mags, ’22)? If so, then let JMM know by sending that juicy news on over to jeff.inman@drake.edu. JMM will treat it like this great workshop for the AI-curious from SJMC’s Chris Snider and computer science’s Chris Porter and tell everyone about it.
Finally, JMM wants to join the Kabin Crew.
"I will show you how to rock that stage."