VRT
Search
2023-09-06 - Fotoshoot VRT MAX - 070.jpeg
VRTInnovation

May 05, 2025

AI and Journalism: A new headline for news

On the occasion of World Press Freedom Day, Karel Degraeve, Innovation Expert at VRT NWS (news department of Flemish public broadcaster VRT) shares his views on the challenges and opportunities of AI for journalism and the media.

AI is a reality that can no longer be ignored, not even in the media. We must prepare for a future where news and journalism will be processed with and by AI, tailored to the individual user. The good news is: journalistic talent will make a big difference here.

Let’s get straight to the point: pessimists see the future of the media as bleak. Media organisations are under pressure, the business model is shaky, and the impact seems to be diminishing.

According to the study, Editorial Media as Defenders of European Democracies, by the leading Scandinavian media company Schibsted, there are several reasons for this. Social media platforms are winning the battle for attention and are accruing most of the advertising revenue. Young people are avoiding news and see information as a free good, and question journalistic habits. Successive financial crises have lead to major cutbacks and thus fewer editorial resources.

As if that weren’t enough, illiberal economic and political forces are waging an information war in which they openly target the ‘mainstream media’. Steve Bannon, strategist for Donald Trump during his first presidential campaign, literally said: “The Democrats don’t matter. The real opposition is the media.” This is also reflected in media policy today, and not just in America.

Finally, Big Tech, the major engines behind digital transformation, seem to favour a strategy of bypassing traditional media. Elon Musk, for example, is working on X Stories, an AI-driven platform that eliminates journalism, or, as he puts it, “puts citizen journalism first.” In short, he wants to use X.ai to detect and generate news stories based on posts on X, and make them personalised in such a way that each individual is seemingly served by their very own ‘newsroom’. Without curation or control, this gives fake news free rein.

A new era

We can all look at this with pity, but that is the wrong attitude. Anyone who has experimented with an AI chatbot for an hour realises that this technology has enormous potential. We are past the hype, we are entering a new era.

This feeling is eagerly shared within the media, and that gives optimists courage. There is, it turns out, a great willingness to innovate among all players, commercial and public. And it is often worked on jointly: knowledge, experiences, and insights are shared in ‘Future Media Hubs’ (an initiative of VRT), at international conferences, and during AI Summits.

The results of this are already tangible in most newsrooms today. Since the launch of the first accessible Large Language Models (LLM), investments have been made in AI-driven tools, preferably developed in-house. The aim is to streamline existing workflows and improve the user experience.

Everyone is convinced that people will play a crucial role … What it comes down to is this: even more investment must be made in journalistic talent.
Karel Degraeve

Smart News Assistant

At VRT NWS, for example, we have developed and rolled out Smart News Assistant, an AI toolbox that assists the newsroom with an ever-widening range of tasks, such as optimising headlines, converting audio to text (and vice versa), and generating summaries. This frees up time and space for our journalists to focus on our core journalistic tasks.

We see even more possibilities with AI in the future. There are plenty of examples that deserve to be followed. Take the small, regional Norwegian newspaper iTromso. Based on conversations with dozens of experienced regional journalists, they developed an AI tool that scans municipal documents for news value. With results: they produced five front-page stories in their first week using this system.

It goes fast

At this stage, the attention is on applications that improve existing operations, rather than drastic changes. Those are happening elsewhere, at the big tech companies: on their platforms, or their smartphones.

For example, the classic Google search is gradually being blown away by the generative search functions of ChatGPT or Perplexity.ai: you ask a question and get a direct AI-generated answer, based on multiple, reliable or unreliable sources. Until recently, this was a text application, but now you can also have a spoken conversation with it. And it sounds increasingly natural.

Growth is rapid. 68 percent of Flemish youth between 18 and 24-years-old use a generative-AI application at least once a month. That is a quarter more than a year ago. And in America, more subscriptions to chatbots are now being sold (13.75% in the 18 to 54 age group) than to news channels (12.25% in the same age group).

“Adapt or die”

That, along with a number of other innovations such as the aforementioned X Stories by Elon Musk, can change the entire news ecosystem. We, as media, will have to adapt to that. The tone today is quite compelling: “adapt or die,” it is said. Or as Gard Steird, editor-in-chief of the Norwegian newspaper VG, puts it: “move fast or go out of business.” The timeframe within which this must happen is short: some speak of 18 to 24 months.

But… what exactly do we need to adapt to? What does the future look like? No one can say for sure today, so Danish futurist Sofie Hvitved advises us to prepare for “possible futures.” Plural, that is.

Let’s get that going.

To give you an idea of such a ‘possible future’: AI agents will scrape the web to generate answers to news questions from individual users. Those answers come in the form desired by the questioner. This will usually be short-form video, or audio (‘voice’ in AI terminology, a voice you can converse with). Websites thus become ghost places: no one comes there anymore, only those AI agents. And oh yes, no one reads anymore, the written article is “almost dead.”

Anyone who has experimented with an AI chatbot for an hour realises that this technology has enormous potential. We are past the hype, we are entering a new era.
Karel Degraeve

The role of journalists

According to the influential journalist and data scientist Nikita Roy, host of the excellent podcast ‘Newsroom Robots’, this is the moment to “reinvent” journalism, redefine our basic principles, and sharpen the news values that drive us. Because with that, especially in a world where distrust is the new normal, we will make the difference.

Every newsroom must redefine those values for itself, and then redesign its organisation and prepare for a future where news and journalism are processed and offered by and with AI. In doing so, says Roy, we must start with a blank slate.

The good news is: everyone is convinced that people will play a crucial role in this. That is, if you think about it, a strange sentence. What it comes down to is this: even more investment must be made in journalistic talent.

A divine being?

There is meanwhile division over how media should deal with Big Tech companies. A fault line is emerging. Large, international media houses are taking the leap forward and making temporary deals with one of the Big Tech companies. Smaller, more national players and public media (such as VRT NWS) do not: they form a front and jointly negotiate with Big Tech.

This is also because, insiders know, Big Tech companies have little or no interest in negotiating with small players. They are ‘too small’, ‘too local’.

The government has an important role to play in this, everyone realises. Policymakers at national and international levels must continue to work on a thoughtful framework that protects editorial media, stimulates innovation, and, as the European Democracy Shield puts it, guarantees a reliable information framework.

Meanwhile, we journalists must continue to do our work, make thoughtful choices, and focus on stories that make a difference. In doing so, we must not fail to critically scrutinise Big Tech companies. On the contrary.

There is a reason why the Vatican studies Silicon Valley so closely, knows American tech journalist Karen Hao: because they really believe they are creating “a divine being with AI.”

Karel DegraeveKarel Degraeve