Deepfake disclosure
Applies from 2026-08-02 · Last reviewed 2026-09-10 · Current
Under Article 50(4) of the EU AI Act, a deepfake is AI-generated or manipulated image, audio or video that resembles existing persons, objects, places or events and would falsely appear authentic or truthful. Deployers must disclose that the content is artificially generated or manipulated. For evidently artistic, satirical or fictional work, disclosure must still happen, but in a way that does not hamper display of the work. The obligation applies from 2 August 2026.
Who is affected
- Agencies producing synthetic spokespersons, avatars or voice clones.
- Social teams altering real footage of people or events.
- Legal reviewers assessing consent and likeness risk.
What triggers the rule
- Content resembles an identifiable person, real place or real event.
- A viewer could reasonably believe the content is authentic.
- Voice cloning or appearance alteration of real people.
- Realistic synthetic scenes presented as recordings.
What does not trigger it / exceptions
- Evidently artistic, satirical or fictional deepfake work: disclosure applies, but the manner can be appropriate to the work.
- Content that is clearly unrealistic and not resembling real persons, objects, places or events falls outside Article 50(4).
Scenario table
| Scenario | Outcome | Why |
|---|---|---|
| AI avatar presenting a product | Disclosure required | Realistic synthetic person could falsely appear authentic; platform labels plus clear wording. |
| De-aged or retouched actor face | Human review required | Alteration of an identifiable person’s appearance — consent and wording must be reviewed by a human. |
| Cartoon-style AI illustration | No disclosure identified | Evidently fictional and unrealistic; not resembling real persons, objects, places or events. |
Required action
- Disclose artificial generation or manipulation clearly and distinguishably.
- Use the platform-native label where available.
- Keep evidence of what was generated, with which tool, and who approved it.
- Obtain consent before using identifiable people’s likeness or voice.
Practical marketing examples
- Synthetic spokesperson reading a script — label the avatar and the synthetic voice.
- AI video “from” a product launch that never happened — disclose the synthetic event.
Relation to platform policy
EU AI Act Article 50(4) requires deployers to disclose realistic AI-generated or manipulated image, audio and video resembling real persons, objects, places or events. Use the platform’s native label and keep your own evidence record.
Run the free check — result, wording and sources in under a minute.
Primary sources
Related pages
Methodology: Disclara answers are generated from a versioned rule set citing primary sources, reviewed by humans. Information on this page is not legal advice. See sources & methodology and rule changelog.