Transformation is a permanent state

AI has arrived in newsrooms – as a tool, an accelerator and a challenge all at once. But as the technical possibilities grow, so does the pressure: How do media outlets remain visible when information increasingly comes directly from chatbots? How do they manage transformation without losing sight of the human element? And why does verification become a competitive advantage in the age of AI-generated fakes?

A conversation with Sven Gösmann (editor-in-chief) and Astrid Maier (deputy editor-in-chief and head of strategy). The questions were posed by editor Ann-Marie Utz.

Ann-Marie Utz: Sven, when you look at our customers’ newsrooms, where do you currently see the most profound changes?


Sven Gösmann: I see two closely related trends. First, the market is consolidating. This is leading to editorial teams being merged and managed more centrally. Second, artificial intelligence is becoming part of daily work. It’s the colleague who never gets tired: a tool that provides support, speeds up rou­tines and frees up capacity. At the same time, AI is also a challenge because it raises new questions about quality, responsibility and control. Together, these two factors are changing our understanding of what journalism must achieve – and how it’s organised.


Ann-Marie Utz: This new understanding likely comes with uncertainty. Astrid, when it comes to customers, where do you currently see the greatest pressure?

Astrid Maier: A key uncertainty concerns the business models. The last few years have already been heavily impacted by transformation: Many outlets have, for example, established premium subscriptions and achieved growth. Now AI is changing search and usage behaviour once again. When people are no longer searching but instead asking chatbots and receiving complete answers, the willingness to click on the linked source behind it decreases. For media brands, this is an existential question: What happens if reach and visibility collapse – and with them, potentially, their digital business model? This leads to the next stage of transformation: How do you stay visible? It’s no longer enough to optimise content for SEO – in the future, it will be more about discoverability and relevance in AI-driven systems. And linked to this is a question of trust: How do media brands strengthen their journalistic profile in an environment where a lot of things look like “answers” but aren’t necessarily verified?

Ann-Marie Utz: Trust and visibility are closely linked to verification. Astrid, what does that mean specifically?

Astrid Maier: It has a lot of practical consequences. It’s about consistently mastering verification and credibly communicating this expertise to the outside world. This requires getting close to people: direct contact – including offline – events, discussions, visible authorship. “Showing your face” means explaining who’s behind the journalistic work, how it’s created and why it’s reliable. I even see some­thing positive in this: Journalism can once again be understood more strongly as a direct relationship with the audience.


Ann-Marie Utz: If AI is changing usage and pro­duction, what else do our customers need for the transformation? More guidance, more speed, more training?

Sven Gösmann: A bit of everything. In my view, the key thing is to provide certainty in an uncertain situation. Security in the sense of “I’m not missing anything” and “What I’m getting is accurate.” This is precisely where our customers – and we at dpa, too – can distinguish ourselves from the noise out there. There’s a lot circulating out there: Some of it is cor­rect, some is driven by interests, some is false. In this environment, trust becomes the decisive differentiator. dpa has built a relationship of trust over decades. This has to be reaffirmed every day – just like in a good relationship: You accompany each other through the day and rely on each other.

Ann-Marie Utz: What does that mean for us internally? How should we approach transformation ourselves?

Sven Gösmann: For me, there are three levels. First: active listening – to customers and shareholders. We need to understand what drives them, where their bottlenecks lie and how we can help in concrete terms. Second: analysing the market, including internationally. Where are platforms, formats and usage habits heading? Third: thinking outside the box and learning from other industries. I often use the music industry as an example. Transformation has changed many things, but not the demand for music and emotion. Live experiences still exist; formats have shifted, and new platforms have brought enormous reach. For us, this means that we have to understand which forms of information delivery and journalistic context will be in demand in the future – and then derive appropriate offerings from that.

Ann-Marie Utz: Astrid, what would you focus on – what is different today compared to the past when we talk about transformation?

Astrid Maier: For me, the most important thing is this: There’s no such thing as “finished.” In the past, projects often had a clear beginning and end. Today – especially with generative AI – the pace of development is so fast that every supposed understanding is immediately followed by the next wave: new AI models, new forms of agents, new interfaces, new possibilities. That demands resilience. And we need to bring people along with us. Transformation rarely fails because of the technology. You can build skills, provide training and hire new talent. But transformation often fails because of the culture, resistance and fear of losing control. Our task – both internally and with customers – is therefore to provide guidance, realistically highlight opportunities, while not downplaying risks, and give everyone the feeling that they are being seen. No one has to tackle transformation alone.

Ann-Marie Utz: A large part of the future seems to revolve around changing target audiences. Many younger people consume news through creators and social media. How can we support our customers in reaching these people?

Astrid Maier: We have strategically shifted towards a “video first” approach. This means that content is no longer conceived solely from a text-based per­spec­tive; instead, video is planned in early on. If you have video, you can derive a lot from it: text, photos, sound bites, social clips. In addition, we have to learn other narrative forms. One example is our clean-feed approach: short videos in portrait or landscape format, without sound, with a script – so that editorial teams can quickly build social media formats from them. I also see dpa as an interface between the new and old worlds: We can help traditional media outlets translate quality journalism into creator logic. And conversely, we can support creators in understand­ing journalistic standards – for example, when it comes to research and fact-checking.

Ann-Marie Utz: That’s exactly where verification comes in. What specific skills do people need today?

Astrid Maier: Anyone working in journalism must know how to verify facts and how to spot AI-generat­ed fakes. It’s no longer a niche topic. It’s a core competency in a world where images, videos and voices are becoming increasingly easy to manipu­late. For us, this represents a major area where we can support customers – both established and new target groups.