Media groups agree on joint guidelines for AI transparency

On the initiative of Media Campus NL, four major Dutch and Belgian media companies — NPO, DPG Media, Mediahuis and Talpa — jointly developed the AI guidelines. The aim is to make clear when AI plays a role in content, without automatically requiring a label for every use of AI.

The guidelines follow the European AI Act, which introduced new transparency requirements for certain forms of AI use in August 2026. For example, providers of AI-generated or manipulated content may be required to clearly disclose that the content was created or altered using AI, particularly in cases such as deepfakes.

The new guidelines are intended to help media companies apply these rules in practice. The focus is not only on legal compliance, but also on when viewers or listeners should actually be made aware that AI has been used.

Not every use of AI requires a label

The guidelines published by the media companies cohort distinguish between different forms of AI use and the context in which content is published. News and journalistic content may be subject to different principles than reality TV, entertainment, fiction or infotainment.

The guidelines also state that not every use of AI needs to be visibly labelled. Technical improvements or applications that do not change the content, meaning or context of a production generally do not require a label.

The guidance provides creators with practical instructions on when content can be considered a deepfake, which AI logo to use and when a visual label or AI audio message is required.

Guidelines are a starting point

Media Campus NL stresses that the joint guidelines are a starting document. The participating media companies intend to further develop the approach in the coming period based on experience and feedback from the sector.

The participating organisations are also free to adopt their own broader transparency policy. This gives media companies room to go beyond the minimum legal requirements.

For media organisations, the challenge ultimately lies in balancing transparency, clarity and proportionality. The joint guidelines are primarily intended as a practical tool: visible when necessary and restrained where possible.

How IO+ uses AI

At IO+, transparency about the use of AI has long been part of our editorial practice. We use AI as a tool within the production process, but a human editor reviews the content before we publish it.

That transparency also has a downside. Mentioning the use of AI can lead readers to judge content differently before they have even read it. For example, identifying ‘Laio’ as an AI writer may give the impression that an article was produced entirely by AI, while a human editor remains responsible for the final text. We call this the ‘transparency paradox’: openness about AI use is important, but an AI label can also trigger negative associations.

That is why we do not see AI at IO+ as a replacement for the editorial team, but as a tool within a process in which human responsibility remains central. This approach aligns with the new sector guidelines, which emphasise that transparency should primarily be informative and proportionate.