Fines Up to €35 Million: EU Introduces Mandatory Labeling for AI Advertising

Affiliate marketers driving traffic to Europe have a new headache to deal with. Starting August 1, 2026, a law on mandatory labeling of content generated by neural networks comes into effect across the EU. While this was previously just a recommendation, failing to label your content can now result in a massive fine. 

What the New Law Demands

Now, any advertisement utilizing neural networks must be clearly labeled:

• AI-generated creatives — videos, banners, and images created via Midjourney, Runway, Sora, and other neural networks.

• AI copywriting — texts written using ChatGPT, Claude, and other large language models.

• Voice actors — voiceovers generated by neural networks. 

• Deepfake content — any video or photo featuring altered faces or voices. 

The label must be noticeable and clear — no tiny fonts in the corner or hidden text strings. The user must immediately understand that they are viewing AI content. 

The fine amounts are strictly defined in the legislation: 

Up to €35 million or 7% of a company\'s annual turnover — for utilizing AI in prohibited areas. For example, behavioral manipulation or hidden surveillance.

Up to €15 million or 3% of annual turnover — for refusing to label deepfakes and AI-generated content, as well as for ignoring ad transparency rules.

Up to €7.5 million or 1% of annual turnover — for submitting fake reports or inaccurate data to regulatory bodies.

Of course, no one is going to issue a €15 million fine for a standard entertaining video or an affiliate campaign setup. For small companies and startups, the law includes a specific rule: their fines are calculated based on the lower threshold, meaning as a percentage of turnover rather than a fixed sum in millions of euros. 

If moderation detects clear signs of generation without a tag, the ad is initially rejected with a requirement to specify the origin of the content. 

Severe sanctions are triggered if the lack of AI labeling is linked to intentional deception or fraud.

• Passing off a generated character as a real, existing doctor in nutra offers.

• Creating fake video reviews with AI faces to promote financial schemes or gambling, imitating real people.

How Affiliate Marketers Can Adapt

For affiliate teams working with Europe, there are several pathways available: 

• Automated labeling. Software developers are updating their scripts so that systems automatically apply a "Generated with AI" watermark onto every single banner. They try to fit this tag neatly into the creative design so moderation passes it, while ensuring users aren\'t scared off and the CTR doesn\'t drop. 

• A hybrid approach. Some media buyers are partially abandoning full neural network generations, shifting instead to working with live source assets where AI only performs secondary functions like background replacements, which do not fall under mandatory labeling rules.  

The initial wave of fines will show just how strictly violations are going to be interpreted. However, one thing is clear right now: the era of unchecked AI usage in advertising for Europe is over.