On 2 August 2026, a new phase of the EU AI Act came into force. Certain advertising content generated or manipulated by artificial intelligence must now be clearly labelled as such. In Belgium, the Conseil de la Publicité is preparing practical guidelines to help advertisers and agencies navigate the change. What it means for your campaigns.
What Article 50 of the AI Act changes
On 20 July 2026, the European AI Office published the final version of its guidelines on transparency obligations. They clarify how to apply Article 50 of the EU's AI regulation, which entered into application for this part on 2 August 2026. The goal: helping the public recognise content generated or manipulated by AI, especially when it could be mistaken for a real image, voice or situation, what the AI Act defines as a "deepfake".
Not every AI touch-up is covered
A colour correction, a crop, or an AI-generated background remain free of any label, as long as they don't mislead the public about the content's authenticity. On the other hand, a virtual influencer claiming to use a product, a synthetic voice imitating a real person, or an entirely fictional product demo presented as real fall under the new obligations. The label must be clear, visible from the first exposure, without requiring a click or hidden metadata.
Belgium is preparing its own guidelines
The Conseil de la Publicité, the sector's self-regulatory body founded in 1974 (through the JEP, the advertising ethics jury), is preparing practical guidelines tailored to the Belgian market with its partners. They are expected within the coming weeks and will complement the 12 principles already established for the ethical use of AI in advertising, ahead of any new regulation.
The AI view, people first
Generating a visual or a voice with AI takes minutes. Deciding whether that content needs a transparency label, wording it without breaking the creative impact, and checking it displays correctly on every platform, Meta, Google, LinkedIn or TikTok, is a matter of expert judgment, not an automatic setting. AI executes. Expertise decides, and watches.
This week's to-do
Review the AI-generated visuals, videos and voices used in your current campaigns. Identify those that could be mistaken for real content, a person, a voice, a product demo. Prepare a clear transparency label while waiting for the Belgian guidelines to be published.
Designing campaign visuals, with or without AI, while anticipating these transparency obligations, is Vistalaro Studio's job. Checking the compliance of your ads on Meta, Google, LinkedIn and TikTok, and adjusting targeting accordingly, falls under Vistalaro Reach. Vistalaro Pilot supports your strategic decisions on the responsible use of AI in your communication.