‘A torrent of crap’: News publishers see trust as a differentiator as AI content proliferates
The rise of generative artificial intelligence is threatening to disrupt some publishers’ business models, but executives from leading global news brands believe it could ultimately benefit their businesses by helping them establish stronger relationships with audiences.
As Google and Microsoft introduce AI responses to search results and new audience-facing tools such as ChatGPT gain traction, it’s becoming increasingly clear that publishers relying on the repackaging of commoditized information could face a difficult future.
But for publishers of original, high-quality journalism and analysis, the proliferation of AI-generated content could present an opportunity to position themselves as essential resources for navigating the AI maelstrom.
Speaking at The International News Media Association World Congress on Thursday, senior executives from publishers including The New York Times, News Corp, Mediahuis, and HT Media predicted that people will see growing value in trustworthy, rigorously produced, and fact-checked information in the coming years.
“AI is almost certainly going to usher in an unprecedented torrent of crap… I suspect you’re going to need to use [publishers’] brands as proxies for trust,” said New York Times publisher A.G. Sulzberger.
Gert Ysebaert, CEO of Belgium-based Mediahuis argued that news publishers must clearly delineate between “content” and “journalism” in order to communicate their value more effectively to consumers – particularly in a situation where much of the content they encounter is generated by AI.
“We’re not in the content business, we’re in journalism. And journalism is about the human view and a human touch – that’s where we will make a difference,“ he said.
The subscription opportunity
Executives at the event said they were hopeful a deluge of AI-generated content would ultimately help them build and strengthen direct paying relationships with their audiences, specifically via subscriptions. If credible and trustworthy information gets harder to come by, audiences will increasingly be willing to pay for access to it on an ongoing basis, the thinking goes.
“The way [publishers] have to differentiate is around trust,” Ysebaert said. “People will be willing to pay if they are engaged and if they trust us.”
HT Media CEO Praveen Someshwar noted that entering into recurring paid relationships with publishers could become more appealing to consumers if AI is employed at scale in an attempt to turbocharge advertising-based business models.
Various ad-supported publishers have now stated their intent to rely – to some degree – on AI content generation, and those announcements have often coincided with layoffs within editorial departments. That could present opportunities for quality publishers orienting their businesses around subscription-first models.
“When there is trust at the highest level, audiences are happy to pay for subscriptions. When there isn’t, that’s when content is monetized through advertising,” Someshwar said.
NYT’s Sulzberger implied the company is confident its existing subscription approach will continue to prove effective in an AI-first world. The deceptively simple “secret” to its subscription growth remains consistent: “It’s about having stuff that’s worth paying for,” he said.
Wikipedia on amphetamines
Despite their optimism, publishers of course remain nervous about the ways AI – and AI-driven search products specifically – could impact their businesses.
Google’s plans to provide AI-generated answers to users’ queries within search results pages could significantly reduce the volume of traffic it sends to third-party sites, for example, which may prompt publishers to attempt to limit the search giant’s access to their content if they feel a value exchange is lacking.
But News Corp CEO Robert Thomson suggested major AI players are well aware of their models’ limitations, and that they’ll continue to value ongoing access to publishers’ content to help them inform timely, high-quality responses in an increasingly competitive generative AI landscape.
“Generative AI is retrospective. It needs to be contemporized with content, and that’s the advantage [news publishers] have,” Thomson said, “Otherwise, it’s just Wikipedia on amphetamines.”
As Toolkits has noted previously, that could result in a situation where publishers are compensated directly for granting AI models a window into their content, and/or indirectly with help growing paid relationships with their audiences. News Corp already struck a multi-year deal with Google in 2021, for example, and The New York Times signed a similar three-year arrangement worth a reported $100 million earlier this year.