Guides · Article 50 · EU AI Act
Disclosing AI-generated images on social media
Posting AI-generated images to social feeds raises a natural question: does the EU AI Act require a label, and if so, whose job is it? Two paragraphs are worth separating: Article 50(2), the provider-side marking duty, and Article 50(4), the deployer-side deepfake disclosure. This guide explains both from the regulation text. It is informational only, not legal advice, and does not name any platform or generator as in or out of scope.
The marking duty is on the provider — Article 50(2)
Article 50(2) asks providers of AI systems generating synthetic audio, image, video or text to ensure the outputs are marked in a machine-readable format and detectable as artificially generated or manipulated. Crucially, this obligation is written on the provider of the generating system, not on every individual who reposts an image to a social feed. So the machine-readable marking question tends to sit with the tool that made the image, and separately with the platforms that build detection and labelling into their products. Whether a given generator falls inside 50(2) is fact-specific.
Deepfakes of real people — Article 50(4)
Where a posted image constitutes a deep fake — for instance a convincing depiction of an identifiable real person — Article 50(4) is written on the deployer of the system to disclose that the content has been artificially generated or manipulated. That deployer-side question can arise for an account or organisation posting such content, and is distinct from the provider-side marking duty. Where the content forms part of an evidently artistic, creative, satirical, fictional or analogous work, the obligation is limited to disclosing the existence of generated content in a manner that does not hamper enjoyment of the work. Whether your post is a 'deep fake' and whether a carve-out applies are fact-specific.
Two separate questions
The practical point is that 'labelling AI images on social' bundles two different obligations on two different parties: the provider's machine-readable marking (50(2)) and the deployer's deepfake disclosure (50(4)). Sorting which, if either, applies to your situation is the useful first step. Our free scope check gives an informational read; confirm the analysis with qualified counsel.
Common questions
If I post an AI image, do I have to label it?
The machine-readable marking duty in Article 50(2) is written on the provider of the generator, not on every poster; the deployer-side deepfake disclosure in 50(4) can apply where the image depicts an identifiable real person convincingly. Which applies to you, if either, is fact-specific — verify with qualified counsel.
See what may apply to your business
Answer seven quick questions for an automated, informational indication of which Article 50 obligations appear likely to apply — free, and not legal advice.