Newsroom leaders will take a more cautious approach to generative AI in 2024

This editorial series examines industry trends across the media, media buying and marketing sectors as 2023 closes and the new year begins. More from the series →

AI,” “ChatGPT” and “hallucinate” were all deemed the “word of the year” in 2023, as advancements in generative artificial intelligence technology swept across the world, in large part due to the launch of OpenAI’s chatbot ChatGPT and followed by efforts from tech giants like Google and Microsoft to develop their own large language models (LLMs).

It wasn’t long before the impact of generative AI on media organizations and their newsrooms became a hot topic. Digiday spoke to four heads of editorial teams — from Business Insider to Trusted Media Brands — to hear how they will approach the technology next year. Most of them said they’re taking a slower and less reactive approach to avoid the public mishaps they’ve seen at other publishers.

Here are their predictions on how the technology will change the media industry next year — and what they’re doing about it:

Integrating AI slowly and internally, while keeping an eye on others’ mistakes

Execs at Business Insider, Forbes and Trusted Media Brands were adamant about the importance of approaching generative AI with caution, and not rushing into using the technology — especially for editorial production.

“As fast as we all kind of want to embrace this tool — and I think there’s a lot of exciting elements to the tool — there’s a lot of tricky elements. There’s legal questions as well as ethical questions,” said Business Insider’s editor-in-chief Nicholas Carlson. “Our posture is, [AI] is going to change things. And so we might as well be proactive and think about it and start working on it. But journalism is one of those professions where you do something wrong, people get hurt. And so we really want to be careful.”

Next year, Carlson said his editorial team will take “incremental steps” to integrate generative AI into their jobs next year, but don’t expect any big changes. “Will it feel revolutionary? No. Will the cumulative effect of those incremental steps in five years feel revolutionary? Yes, if you compare them over time,” Carlson said. “We got to put one foot in front of the other and not blow up our reputation,” he added.

Jacob Salamon, TMB’s vp of business development, said he’s been keeping track of the “cautionary tales” of what happens when publishers rush into using the technology too quickly and make “silly mistakes.” Salamon’s focusing on internal applications for the technology, such as how it can help writers with their first drafts and research, “before we think about anything that’s going to be customer-facing or reader-facing. Because that’s way too risky. [AI is] way too new to kind of dive all in on that.”

The technology will continue to improve — and get ‘scarier’

Forbes’ chief content officer Randall Lane was bullish on getting his employees’ hands on generative AI tools so that they could familiarize themselves with the capabilities as the technology continues to improve and develop and become more integral to different job functions. And he’s not alone. A number of the task forces publishers created earlier this year to oversee generative AI initiatives now involve larger groups testing the tech.

“We want to be as proficient as possible in this technology, which is the most important technology over the last 20 years… It’s not going anywhere,” Lane said. “Every three months, it’s another step change.”

But Salamon expects the generative AI tools to get “scarier” as the technology becomes more advanced next year. (OpenAI, which made its latest LLM GPT-4 available to developers in July, has already filed trademark applications for “GPT-5” in July and in October filed for “GPT-6” and “GPT-7″.)

However, Carlson said using the technology will soon be a normalized part of working in the media. “All this drama about using it at all, it’s just going to quickly fade and it’ll become a regular, normal part of most journalists’ lives in terms of being a productivity tool,” Carlson said. “The technology is going to improve. We’ll wake up one day and it will really be a core part of what we do today as journalists.”

AI won’t be able to replace journalists

Some publishers’ experiments with generative AI has led to what can only be called public embarrassment. Recent examples include a report from Futurism alleging that Sports Illustrated had published articles with fake author names and headshots from an AI-generated image website (and just weeks later, three execs — including SI’s CEO — were fired). Gannett faced backlash and eventually paused its experiments with LedeAI in August to write local high school sports stories, after several errors were mocked on social media.

These mistakes are evidence that AI won’t be able to replace journalists, execs told Digiday.

“We see AI as a tool for our journalists. It’s not — nor will it ever be — a replacement for journalists. It’s a way — like the internet itself — to make journalists better,” Lane said. “The path we’re on… [is] always remembering that AI is not going anywhere, it’s here to stay and that it’s a tool to make our journalists stronger, not anything that’s supposed to be any semblance of what a reporter is.”

Carlson echoed this: “Journalism will always require what it requires now, which is a human to sort of say, ‘Yes, this is true and we stand by it.’ I see other people who don’t think that. I think they’re crazy.”

Newsroom unions have demanded their employers this year agree to new terms on how generative artificial intelligence is used and the impact it will have on their employees to protect their jobs from being threatened by the technology.

But some publishers are fully leaning in

Ingenio’s Jaffe believes the impact of generative AI will be significant enough to give digital publishers “the opportunity to reinvent their businesses and reignite growth” and is embracing the technology with open arms.

He believes chat interfaces are the future of content consumption, and that gives publishers the opportunity to create more engaging and personalized experiences for readers on O&O properties. (On average, users are spending more than 10 minutes with Ingenio’s spiritual guide chatbots.)

“It’s something that can rekindle growth, something that’s incredibly powerful for us to deliver better experiences for users and it’s coming at the right time that we really need it, given this year’s confluence of traffic declines from Google and Facebook, decreasing ad rates, audiences’ migration to short-form video and the absorption of publishers’ content into LLMs without receiving payment, he said. 

“It’s been a tough [year] and I shouldn’t be as excited about next year as I am based on all of these challenges, but the innovation around generative AI gives me a lot of hope,” Jaffe said.

More in Media

Media Briefing: Publishers reassess Privacy Sandbox plans following Google’s cookie deprecation reversal  

Google’s announcement on Monday to reverse its plans to fully deprecate third-party cookies from its Chrome browser seems to have, in turn, reversed some publishers’ stances on the Privacy Sandbox. 

Why Google’s cookie deprecation reversal isn’t actually a reprieve for publishers

Publishers are keeping a “business as usual” approach to testing cookieless alternatives despite Google’s announcement that it won’t be fully deprecating third-party cookies after all.

Immediate deepens CMP strategy, slashes ad tech partnerships for sharper data governance

Consent management platforms at Immediate aren’t just about ticking boxes for data laws.