Is OpenAI Quietly Building a Media Content Empire?

OpenAI has proven to be a magnet for lawsuits from authors and publications who say the company illegally scraped their work from the open internet to train the large language model behind its hit chatbot, ChatGPT.

But at the same time, the company has been amassing a trove of legally sourced content through dozens of media partnerships -- most recently teaming up with magazine giant Condé Nast. In all, OpenAI has forged 31 such partnerships in the last two years, including high-profile deals with Time, The Atlantic, the Associated Press, and many others.

The internet boom of the 1990s saw media advertising models suffer as advertisers left traditional media outlets like newspapers and magazines in droves. That trend has continued in recent years. Statista shows global print ad spending dropped from $62.02 billion in 2017 to $35.85 billion in 2023 and losses are projected to fall by more than 50% by 2029. Media companies are hungry to replace that lost revenue. And GenAI’s need for quality content may give those companies a real boost.

“Over the last decade, news and digital media have faced steep challenges as many technology companies eroded publishers’ ability to monetize content, most recently with traditional search,” Condé Nast CEO Roger Lynch told employees in a memo. “Our partnership with OpenAI begins to make up for some of that revenue, allowing us to continue to protect and invest in our journalism and creative endeavors.”

Related:More Newspapers Join the Copyright Battle Against OpenAI, Microsoft

Other media executives cite similar goals.

During an Aug. 10 earnings call, News Corp CEO Robert Thomson said his company’s partnership with OpenAI offers guaranteed revenue. “It’s tangible, it’s meaningful,” he said. “The way to regard the OpenAI agreement is that there will be a guaranteed amount and the two companies will be working to create products that will generate even more value for both companies.”

The Condé Nast partnership will give OpenAI access to a massive trove of content from some of the best-known magazine titles in the world, including its flagship products Vogue, The New Yorker, Bon Appetit, Vanity Fair, Wired, and more.

While official agreements for the partnerships have not been publicly released, ChatGPT itself thinks large sums could be involved. When prompted to answer a question about how much the deal between OpenAI and AP could be worth, the chatbot answered, “For large-scale data licensing involving extensive archives like AP’s, which has a vast repository of news content, the deal could potentially be worth several million dollars, especially given the value of high-quality, reliable date in training AI models.”

Related:Could Google Lose ‘Verb’ Status After Antitrust Ruling?

It gave a similar answer when prompted to provide a value for the OpenAI partnership with Condé Nast.

What This Means for SearchGPT, ChatGPT

After the meteoric success of ChatGPT, expectations were high after the company announced its prototype search engine product, SearchGPT, in July. With Google commanding a 91% market share in the search engine space, OpenAI will have to contend with a well-established competitor and market leader. But after a federal court ruled against Google earlier this month in an antitrust case, the search giant may have to loosen its search engine market grip.

OpenAI is banking on its media partnerships to differentiate its search product with higher quality content, real-time information, specialized expertise, and diverse content. Instead of a simple query, users will be able to converse with the search engine, asking follow-up questions for more clarity and insight.

Jim Yu, founder and executive chairman of enterprise SEO and content platform company BrightEdge, says SearchGPT’s strength is that it can generate a multimodal, comprehensive search result that is conversational. On the plus side for publishers, a new interface engine also brings new opportunities to make money.

Related:Google AI Overviews and Its Business Impact in the Tech Industry

“The world order of search for the last 20 to 30 years has been about search engines’ quality of information, and it’s been a bi-directional relationship,” Yu tells InformationWeek in an interview. “They get content from the publishers and then they flow clicks back into the ecosystem … That economic system is being disrupted. OpenAI has an opportunity to have a very different relationship with the ecosystem.”

Manoj Saxena, founder and executive chairman of Responsible AI Institute, says the partnerships open new doors for revenue generation on both sides. “I think there is a massive opportunity for revenue,” Saxena says. “These companies create content and with a partnership with OpenAI, it’s a great way for them to start getting some credit and start getting some value -- but that’s just the starting piece.”

In return, OpenAI gets better content with more access to expert input. “What the LLMs have done is trained on all of the internet stuff. And who creates most of the internet stuff? The average person -- so whatever elements are there today are representations of the average person in the world. That was just the first level … the second level is superhuman level, and third level is ultrahuman.”

With models trained on better content, the quality of output improves, Saxena says. With expert articles in niche publications and a deep catalog of intellectual content, GenAI models could move to the next levels of intelligence. “To me, this is just the first party of those three stages,” he says. “The nature of artificial intelligence content is going to get very enriching, very engaging, and very intimate.”

Pros and Cons for the Media Industry

While publishers will likely jump at the chance for new revenue streams, the creative talent behind the content may be trepidatious. Unions have already started pushing for worker protections around the use of AI in the newsroom.

After New York Magazine parent Vox Media entered an agreement with OpenAI in May, the magazine’s union released a statement of concern criticizing the lack of transparency around the deal.

“We know from our colleagues and from the reaction to this agreement, that AI continues to be a critical concern for our members, and our industry,” the statement said. “We are frustrated that Vox Media made this deal while actively avoiding any engagement with our proposals on AI at the bargaining table.”

Alex Mahadevan, director of Poynter’s digital media literacy project, MediaWise, says part of the problem is that not much has been publicly released about the deals. “They are getting a ton of content and we really don’t know how much that is worth,” he says. “You are giving all of this [intellectual property] to AI companies to train on. It’s hard to quantify, to know, if this is a bad deal or not. But who else is going to pay for our back archives?”

Mahadevan says while there’s trepidation about AI within media organizations, there’s also an opportunity to enhance journalists’ data gathering abilities. But the benefits should be weighed against the existential risks to creatives.

“I think unions should be concerned about publishers getting too cozy with the AI companies because it is very tantalizing to try to use these tools to shrink the amount that you’re spending on reporting,” he warns. “On the flipside, they can look at it as a positive … more money coming into the organization hopefully means we can start building the newsrooms back up that have been declining.”

For BrightEdge’s Yu, publishers stand to gain profits in the search engine market that they haven’t had access to with the old model of paid clicks and search engine optimization. “There haven’t been new equations for publishers in terms of economic models for search that have been very disruptive,” Yu says. “So, I think this is a pretty interesting opportunity for that broader ecosystem.”

And Yu thinks it’s very likely that other AI companies will follow OpenAI’s lead with media partnerships.