‘Brilliant storytelling needs humans, but AI’s support has benefits’

The European Broadcasting Union’s News Report 2025 explores the real-world challenges newsroom leaders face when adopting generative AI tools, writes Journalism AI. Based on interviews with 20 industry leaders and researchers, it offers practical use cases, insights on staff engagement, audience reactions, and the impact of AI on creativity. Alongside hands-on guidance, the report also considers how journalism can thrive in an increasingly complex information landscape.

The European Broadcasting Union (EBU) writes that media organisations expect a lot from generative AI – particularly in their quest to reach broader audiences and make news work more efficient. But while the tech industry has churned out new tools at speed, newsrooms have become increasingly strategic about developing and deploying AI-supported solutions.

Not only do they navigate accuracy challenges that are tied to public trust. But many of them have also realised that they need to put their journalism and their audience connections front and centre again to maintain visibility and legitimacy when automated content production becomes abundant

This report takes a closer look at the challenges newsroom leaders face when implementing generative AI-based tools. It presents a variety of practical use cases and discusses how to bring staff along, how audiences react, and how the technology might affect creativity.

The author, Dr Alexandra Borchardt, interviewed 20 newsroom leaders who are at the forefront of driving AI-implementation and top researchers in the field. In addition to practical advice, the report provides an outlook for how journalism could flourish in an increasingly complex information environment.

Here is the EBU’s Executive Summary of the report:

The development of generative AI-driven technologies since the publication of the previous EBU News Report has been breathtaking. Still, most of the news industry has moved cautiously, busy researching and testing which features could make news production more efficient, reach broader audiences and add real value to journalism.

The following observations focus on the experiences of newsroom leaders and address the big picture.

Perceptions from the newsroom:

• Newsrooms increasingly use AI tools for improving internal processes through custom GPTs or features built into their content management system (CMS) to enhance performance – but also for developing public facing formats like chat-style queries about news content, personalising the news experience and soliciting and/or structuring user contributions
• Newsroom leaders are generally quite satisfied with using generative AI tools for  everyday tasks that don’t require perfect results, like translation, transcription and subtitling. They have observed massive quality improvements and feel they can serve audiences better with less effort
• Newsroom leaders are hesitant to implement more sophisticated audience-facing products because of prevailing accuracy problems. There are technical fixes for some of these which require additional resources

Newsrooms still consider having a human in the loop’ necessary, perhaps even an asset. It does however constrain the scaling at speed that AI promises

• Newsroom leaders are generally happy with the extent to which staff is interested in AI, and ready to experiment with and use AI tools. Initial fears about digital divides have not materialised. But there are worries about laziness and hidden usage of certain AI tools
• Others think newsrooms worry too much about efficiency improvements and invest too little in equipping journalists to tell better stories with the support of AI

The big picture:

• There is a mismatch between the speed of progress by tech providers and the capacities of media organisations to determine the potential usefulness and risks associated with new models and tools. Evaluation, testing, and implementation take time
• There is still very little cost-benefit evaluation and impact measurement when media companies use AI. Clear KPIs are missing. Also, calculating risks is difficult, since the business models and price policies of tech providers are unclear, but dependencies are huge

Audiences around the world have taken to using AI tools, particularly people in urban areas and in the student population. This means that media organisations not only have an important role in educating the public about AI, but they also need to make a strong case for the importance of quality journalism to society and democracy

• The big issues about data usage, copyright claims and participation in potential economic gains of the AI economy have remained largely unsolved, as have concerns about resource and energy usage. Media leaders consider it essential for their organisations and the tech industry to co-operate if the goal is a healthy information environment
• While misinformation – with or without the contribution of AI – is acknowledged as a significant challenge for media organisations, AI-supported verification tools and techniques are also becoming more sophisticated. Some media leaders are worried about deep fakes on a smaller, individual scale that don’t receive media attention and can still destroy lives

For journalism to survive and flourish in the age of generative AI it has to provide additional value, for example with investigations, brilliant storytelling, uniquely human features like voice and personality, accountability and community connection

• It will be essential for media organisations to be a trusted partner of the audiences they serve
• Read the full report here
• Listen to the Eurovision News Podcast for insights from the report with Borchardt and Olle Zachrison, Head of Artificial Intelligence and News Strategy at Swedish Radio
• Join EBU for a free webinar on 30 April at 11am CET on the report with Borchardt and Belén López Garrido, Editorial Manager at Eurovision News