AI Act: labelling AI-generated content — what changes on 2 August for your campaigns

On 10 June 2026, the European Commission published its Code of Practice on marking and labelling AI-generated content. It prepares the way for the Article 50 transparency obligations of the AI Act, which take effect on 2 August. For Belgian SMEs that use generative AI in their visuals and campaigns, here is what matters.

1. A deadline with a date: 2 August 2026

The Article 50 transparency obligations apply from 2 August 2026. The Code published on 10 June is not a new law: it is an operational, voluntary playbook showing how to comply in practice. In short, the framework is taking shape and the countdown has begun.

2. What "transparency" means in practice

Two layers coexist. For providers of generative AI tools: outputs — image, video, audio, text — must be marked in a machine-readable format (metadata, watermarking) and detectable as artificial. For professional users — you, when you publish a creative generated or edited by AI: clear disclosure to the public is expected in several cases, notably content that could mislead.

3. Why it directly concerns your visuals and ads

In Belgium, nearly one company in four already uses AI, and AI-generated ad creatives are multiplying. A Meta post, a banner, a TikTok video produced with a generative tool falls within scope. The good news: labelling is not a creative brake. Built into your production chain, it becomes a reflex — and a mark of seriousness towards your audience.

The AI lens — humans first

AI is precisely the tool creating the content that needs labelling: so compliance is decided in production. An assistant can apply the right markings, check metadata, flag an unlabelled creative before publication. But AI executes — it does not decide. Judging when a disclosure is required, weighing a campaign's impact against the clarity owed to the public, remains expert work. That is our stance: humans first, AI in support.