How AI is changing work
Journalism & news
Working things up faster, checking more carefully.
Newsrooms mainly use AI at the front end of the work: transcribing, subtitling, searching archives and spotting what could be news. The agreement in the sector is clear: a human checks and decides. Transparency about AI is becoming a standard part of the job.
What changes
- Interviews and council meetings are transcribed automatically.
- Subtitling and reading aloud are increasingly done with speech technology.
- Archives and news feeds are searched for what is relevant to a title or region.
- Articles are converted into other forms, such as audio or another language.
Where people make the difference
- Deciding what is news and which angle is right.
- Verification: nothing from AI is published without human checking.
- Accountability to the public, including about how AI was used.
How AI is already used
Human, machine, human
ANP, the Dutch national news agency, uses AI for ideas, headlines and writing up interviews. Nothing from AI is used without a human checking it.
VillamediaSubtitling with speech recognition
At NOS, speech recognition converts live speech into subtitles. NPO is experimenting with fully automatic subtitling on NPO Politiek & Nieuws.
NOSCouncil meetings transcribed
Omroep Flevoland has council meetings transcribed; RTV Utrecht uses AI to select regional stories for social media. The journalist decides the angle.
SpreekbuisNews read by an AI voice
At weekends, RTV Drenthe has the regional news read by an AI voice; RTV Oost uses a voice trained on one of its own staff.
Broadcast MagazineArticles as an English podcast
De Limburger automatically converts articles into English-language podcasts to reach expats and international students.
VillamediaAgents that prepare news
Mediahuis is experimenting with AI agents that spot news based on a title's editorial formula and write a first version, with a human doing the final check.
VillamediaAn ethical compass for AI projects
DPG Media offers a game format that lets media teams work through ethical questions around AI and data projects, free for the Dutch and Belgian media sector.
Adformatie
Where the sector is heading
NPO, DPG Media, Mediahuis and Talpa jointly drew up guidelines for when AI use gets a label. Technical improvements that do not change content and meaning do not, in principle, need a label.
NPOPeople also get news through AI chatbots: in 2025, one in twenty Dutch people did so, and 11 percent of 18- to 34-year-olds.
Commissariaat voor de MediaIn Flanders, the Raad voor de Journalistiek (Flemish press council) has laid down that the editor-in-chief remains responsible and that newsrooms are open about automated news production.
Raad voor de Journalistiek
What a maker can do for you here
- “Build a workflow that transcribes and summarises our interviews and meetings for the newsroom.”
- “Convert our articles into audio or another language, with a review step.”
- “Make our archive searchable by people, topics and places.”
- “Help us make a decision tree for when we label AI use.”
Researched on 5 October 2026. Know a good example from the Netherlands? Let us know.