Litigate or License? News Publishers Struggle With Letting AI Have Their Content

News publishers are grappling with how to respond to Big Tech’s aggressive push into generative AI models that leverage decades of their news content without paying a dime.

For many, the decision comes down to whether to fight AI purveyors in court, or license their content for a price now — which may seem like a pittance in years to come as Big Tech reaps untold billions from building out the technology.

Tech companies like OpenAI, which has already hit $2 billion in annual revenue, continue to court publishers with up-front compensation  — after having likely used much of that same content to train their Large Language Models (LLMs). A flurry of big-ticket partnerships have been announced in recent months including with publishers like News Corp, Dotdash Meredith, The Financial Times, The Associated Press and Axel Springer. The multiyear deal between OpenAI and News Corp. is worth over $250 million, The Wall Street Journal reported, the largest deal made by the AI firm thus far.

OpenAI has set the pace, with 24 confirmed licensing deals between platforms and news publishers, according to CJR/Tow Center. Microsoft, which also has a partnership with OpenAI, follows with seven confirmed deals and Google has closed on three partnerships. Other companies like Apple, Amazon and Adobe have also inked deals of some kind or are involved in ongoing discussions with publishers, according to Tow.

Some media figures are sounding the alarm on the pace of AI dealmaking and expressing concern over a sense of déjà vu. On Friday, The Information founder and CEO Jessica Lessin penned an op-ed in The Atlantic calling the partnerships a “fatal error” for publishers.

It is “extraordinarily distressing” to observe media organizations rushing to partner with AI firms after being “burned” by tech companies in the past, she wrote. The AI deals are “far worse for publishers” than past deals with giants like Facebook, because it will limit their search traffic in the long term. Even if OpenAI executives ”were ardent believers in the importance of news, helping journalism wouldn’t be on their long-term priority list,” she said.

OpenAI Chief Executive Officer Sam Altman speaks during an event at Keio University on June 12, 2023 in Tokyo, Japan. (Photo by Tomohiro Ohsumi/Getty Images)
OpenAI Chief Executive Officer Sam Altman speaks during an event at Keio University on June 12, 2023 in Tokyo, Japan. (Photo by Tomohiro Ohsumi/Getty Images)

Industry financial strain spurs dealmaking

The push by Big Tech to lock up content to feed its chatbots comes at a difficult time for the news media industry. Media companies have struggled with falling digital advertising revenue — which is now dominated by Big Tech companies — and struggled to maintain subscribers, leading to a wave of mass layoffs and other draconian cost-cutting measures. Just Thursday The Washington Post reported a 50% decline in its web traffic, a $77 million loss in 2023 and looming cutbacks.

The intense financial pressure means media companies are tempted by the prospect of incorporating AI in their newsrooms but also conflicted by what it could mean for the value of their content. At the same time, Big Tech companies have told investors that they plan to spend tens of billions of dollars a year to develop AI and are in an all-out race to gain a foothold in what is predicted to be a multi-trillion-dollar industry.

The licensing deals between media organizations and OpenAI so far have been conducted on an individual basis, with each company granting the AI firm varying degrees of access to content. The AP’s partnership, which is valued in the single-digit millions a year in compensation, according to WSJ, is limited to the use of archival content for training LLMs.

News Corp.’s deal, which is much larger and more comprehensive, allows OpenAI to display content from the organization’s portfolio of mastheads in response to ChatGPT questions from users. The intention is to “enhance” OpenAI’s product offerings, “with the ultimate objective of providing people the ability to make informed choices based on reliable information,” the company said in a release. Alongside the licensing agreement, News Corp. will provide “journalistic expertise” to OpenAI in an effort to ensure quality control of the tech company’s journalistic standards.

Experts say that while some media companies are eager to embrace AI technology and welcome revenue from cash-flush tech giants, it is impossible to ignore the similarities to previous cycles of tech development. Over the past decade, news publishers molded their business and editorial strategies around tech initiatives, including chasing search functions and social features, only to see the tech companies reap the real benefits — as with Google — or suddenly turn off the flow of cash when they pivoted strategy, as with Facebook.

Publishers today are more jaded about the state of news as a business, after years of dealing with companies like Meta and Google — whose constant platform changes, algorithm fluctuations, and updates tend to come at the expense of the media organizations. To avoid the risk of missing out on a diversified revenue stream in a brutal advertising market, many news publishers are simply opting to meet with AI executives, hoping to see the advantages of being non-confrontational over copyright infringement concerns.

“The journalism industry is grappling with testing these tools and making sure that they adhere to journalistic standards,” Hilke Schellmann, assistant professor of journalism at New York University told TheWrap.

Every industry is attempting to determine where AI fits within their specific field — by testing the accuracy and limitations of the technology, or by experiencing the general costs of fundamental industry change. Journalism is no different, and is “on par as any other industry,” said Schellmann, who authored the book “The Algorithm: How AI Decides Who Gets Hired, Monitored, Promoted, and Fired and Why We Need to Fight Back Now.”

Lawsuits building steam

Meanwhile, some publishers, most notably The New York Times, have launched lawsuits against AI developers, hoping to set a precedent for copyright infringement by AI content scrapers of news content. At the least, the news organizations want to be compensated for content that may have already been used without permission.

Eight Alden Global-owned newspapers including the Chicago Tribune have collectively sued OpenAI and Microsoft, with two separate suits also coming from digital media outlets that include The Intercept, Alternet and Raw Story.

AI companies’ primary defense will almost certainly involve claiming fair use, an exception to copyright law, Saurabh Vishnubhakat, Cardozo Law Professor specializing in intellectual property law, told TheWrap. The application of the principle in relation to the training of AI models “is still a new and largely untested legal theory,” with actual court decisions remaining relatively “scarce.”

“It’s definitely copyrighted material that they’ve used for building most of their models,” Schellman said. “If it’s not going to be illegal at the end of the day, I don’t know what is. But I’m not a judge, a judge will decide that.”

Over the past decade, news publishers molded their business and editorial strategies around tech initiatives, including chasing search functions and social features, only to see the tech companies reap the real benefits .

Whether OpenAI and other companies can successfully defend themselves against copyright lawsuits is “unlikely to be clear for at least a few years,” Vishnubhakat added.

The lengthy legal challenges come with a hefty price tag for the publishing companies. The Times is already spending $1 million on fees related to the lawsuit against OpenAI and Microsoft, the company said during its first-quarter earnings.

A.G. Sulzberger
A.G. Sulzberger, Publisher, The New York Times, speaks at the New York Times DealBook conference in 2018 in New York City (Photo by Stephanie Keith/Getty Images)

Meanwhile OpenAI hit a $2 billion revenue milestone in December, according to the Financial Times, with the company aiming to double this figure in 2025.

In April, Microsoft reported revenue of $61.9 billion, up from $52.9 billion a year earlier. Profit hit $21.9 billion, up from $18.3 billion.

Schellman said news publishers and AI firms have the option to settle the lawsuits, avoiding a lengthy battle in court. “If it was clear that my tech company infringed copyright, I probably would settle, and then hopefully, all that content that we used would be free and clear,” he said.

Such a settlement would likely entail OpenAI compensating the news organizations for the previously used content, and more importantly it could avert a public and costly trial. But it could potentially “alter OpenAI’s business model, perhaps significantly,” Vishnubhakat said.

TheWrap reached out to OpenAI and Microsoft who referred us to their partnership announcements but would not otherwise comment. TheWrap also reached out to the New York Times, News Corp, the AP, Axel Springer and the Financial Times. Each referred us back to their previously published statements on AI.

The post Litigate or License? News Publishers Struggle With Letting AI Have Their Content appeared first on TheWrap.