How far does the artistic carve-out go?

There's a persistent belief that artistic or satirical AI content is exempt from the EU AI Act's labelling rules. The actual rule is narrower and worth stating precisely, because relying on the folk version is how commercial content ends up unlabelled.

For deepfake-type content (realistic AI image, audio, or video) that is evidently part of an artistic, creative, satirical, fictional or analogous work, the Article 50(4) disclosure duty is attenuated, not waived: you may disclose in a non-intrusive way that doesn't spoil the work (think credits, a caption, programme notes) rather than stamping an icon across the piece. The duty to disclose the content's artificial origin remains; only its intrusiveness is relaxed.

The boundary that matters commercially: the carve-out does not cover content that is primarily commercial or informative. An ad is not satire because it's funny; a brand's AI-generated "mockumentary" campaign is still primarily commercial; realistic AI imagery in content marketing is informative-commercial, not artistic, however creative the execution. For all of those, the full visible-labelling convention applies: the harmonised "AI" mark at every exposure for images, opening disclaimer plus persistent icon for video, spoken disclaimer for audio.

For genuinely artistic works, the practical pattern is easy: put the AI disclosure where that art form's credits naturally live (end credits for film, the description or liner notes for audio works, the caption or plaque for exhibited images), so it's findable without defacing the work.

Satire aimed at real people deserves a specific caution: the attenuated duty still applies (the audience must be able to learn it's synthetic), and other law (defamation, personality rights) operates entirely outside the AI Act. "It's satire" answers the intrusiveness question, not the disclosure question, and none of the other legal questions at all.