Get weekly AI video rankings, evaluation updates, and industry news delivered to your inbox.
New transparency rules put the compliance burden on deployers, not just AI developers
A photorealistic image of a product, generated by an AI system and dropped into a digital ad campaign, now requires a disclosure label in the European Union. That requirement took effect on August 2, 2026, under the EU AI Act's transparency obligations for deployers.
The rules apply to any company using AI systems to produce or manipulate images, audio, or video for professional purposes. Audiences should be able to distinguish factual content from AI fabrication, particularly when the output is realistic enough to deceive.
Who bears the obligation
The compliance burden falls on deployers, the businesses that use AI tools to produce advertising assets, rather than on the developers who build those tools. That distinction matters for the advertising supply chain, where agencies, brands, and production houses routinely use third-party AI platforms to produce visual and audio content.
The rules extend to AI-generated press releases alongside advertising assets. A company issuing a photorealistic AI-produced image alongside a product announcement would need to label both.
What compliance looks like
The Act requires that AI-produced content be clearly labeled as artificially generated, though the specific format of that disclosure is not prescribed in the sources available. The obligation attaches to content that could plausibly be mistaken for real imagery, audio, or video, the deepfake category broadly defined.
Non-compliance carries steep financial exposure. The EU AI Act's penalty structure is among the harshest applied to technology companies in the bloc, though the exact fine thresholds for this provision are not detailed in the available sourcing.
The advertising supply chain problem
A single campaign can pass through a brand, a production agency, a media buyer, and a distribution platform before reaching a consumer. The deployer obligation means the entity that runs the AI system bears primary responsibility, but in a multi-party chain, that accountability can become contested.
Companies that have built workflows around AI-generated imagery, voiceovers, or synthetic video will need to audit where those tools sit in their production process and verify that disclosure travels with the content through distribution.
The next checkpoint for the industry is enforcement action by EU member state authorities, which can now open investigations into non-compliant campaigns for content published on or after August 2, 2026.
Get weekly AI video rankings, evaluation updates, and industry news delivered to your inbox.