News companies face new questions, challenges, opportunities with GenAI

By John Horstmann

City, University of London

London, United Kingdom

Connect      

Embracing technology safely and ethically “offers new possibilities for all of us,” Katharina Neubert, vice president, strategy and investments at Germany’s Holtzbrinck Publishing Group, said during the INMA World Congress of News Media in London. 

During a workshop on Generative AI, multiple experts provided their thoughts on the possibilities, uses, and threats of GenAI in media.

“AI is bringing a new wave of technology and transformation to humankind and we will and can decide how we approach that,” Neubert continued, emphasising the human factor in the use of AI.

While she noted several uses for GenAI within media companies — including transcribing, creating videos based on text, search engines, chatbots, and customer service calls — she also pointed out several unresolved issues and problems with the new technology.

GenAI has many promising features and uses but still has several unresolved issues.
GenAI has many promising features and uses but still has several unresolved issues.

For example, AI can be biased and is only as qualitatively good as the training data itself, which is derived from human work. She also suggested AI is “intransparent and unexplainable” regarding how it comes up with its results, which can lead to copyright and legal issues. 

In addition, she flagged the concern that the use of AI alone can burn through as much electricity in a year as the country of Ireland.

Despite these issues, she is convinced GenAI is part of the future for media organisations: “It’s not a question of if but how we will utilise GenAI,” she said.

Discovering new AI use cases 

Lyn-Yi Chung, deputy chief editor at CNA Digital in Singapore, said the company has been testing AI for various purposes.

Its use in automated video editing has been particularly successful: “This project is less than a year old but we’re already seeing so much progress. [AI] can lay out a package with pictures over a voiceover. We are really happy with that.” 

She has also been testing the technology for live video transcription, auto-clipping news bulletins, generating summaries for articles, translation, and text-to-speech projects.

But not all projects have gone as planned.

“Be prepared for pushback,” she said. “It’s a long road ahead. AI projects are a lot of work, a lot of time and effort, and a lot of pain. You have to agree on an acceptable success rate. For me, 85% is good.”

She noted that while AI offers many ways to improve the work of journalists and make their lives easier, it’s important users don’t find a solution to a problem that becomes a problem itself. 

“Make AI work for you, not you work for the AI,” she told the audience. 

AI transforms the newsroom

Gina Chua, executive editor of Semafor, and Noreen Gillespie, journalism director at Democracy Forward at Microsoft, discussed how AI is transforming the job of reporting. 

“We have to think about our specific uses [for AI] that we have to see what LLMs and AI can do for us and then map it,” Chua said.

Semafor is using AI company Miso to search for specific articles and summarise them, she said: “It’s supposed to search for analysis, not news. It doesn’t always do that so well. But it’s a very fast way to search. It doesn’t substitute for human judgement, but it speeds everything up.” 

Mike Butcher, editor at large at Techcrunch in the U.K., said he believes every media organisation should be looking into potential uses for AI. 

“I would advise everyone to have a little section of a department working on monitoring this,” he said, encouraging all media businesses to “operate like an entrepreneurial start-up,” when it comes to AI. 

The capabilities of AI will only exponentially improve in the future, he said: “GPT was a ‘wow moment.’ GPT4 is very good. But GPT5 and GPT6 won’t just be a bit better. They will be exponentially better. That’s how this technology works.” 

His main concern with GenAI is its lack of respect for copyright: “OpenAI basically sucked in the entire Internet and all of our stuff is in there. If we can prove that [our] data is inside your large language model, that’s our copyright. 

“Everyone is getting lawyered up about what can be done.” 

Keeping the big picture in mind

Dmitry Shishkin, chief executive officer at Switzerland’s Ringier Media International, gently pushed back on the media’s prioritisation of AI. 

“Please remember that this is not only about AI. This is still about journalism. Our reporting is what will make us stand out,” he said. “If you don’t think about journalism and media, then you don’t even need to bother thinking about business models.”

Regardless the technology, real transformation only occurs when cultures change.
Regardless the technology, real transformation only occurs when cultures change.
 

He said media organisations “... need to remember who we are and why we exist. Media organisations will only survive if they play a significant role in people’s lives. If they don’t, they will struggle.”  

Shishkin emphasised the principles of human connection, originality, and prioritising user needs.

He asked the audience to consider the following questions:

  • What sets you apart?
  • Does your brand have a reason to exist in a future world?
  • Why do you exist in the market?
  • What is your place?

“[To be successful], we need to remember who we are and why we exist,” he said.

About John Horstmann

By continuing to browse or by clicking “ACCEPT,” you agree to the storing of cookies on your device to enhance your site experience. To learn more about how we use cookies, please see our privacy policy.
x

I ACCEPT