Usually — and often under more than one rule. An AI avatar can be the interface people talk to (50(1)), a piece of generated video or audio (50(2)), and a realistic likeness that could pass as a real person (50(4)). Which obligations apply depends on what your avatar does and how real it looks.
Article 3(60) defines a deep fake as AI-generated or manipulated image, audio or video content that resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful.
A digital replica of a real person clearly qualifies. Invented people are less clear-cut. According to the widely reported interpretation of the European Commission's final Article 50 guidelines, adopted in July 2026 — reported by multiple law firms including Bird & Bird, Reed Smith and Davis+Gilbert — the guidelines read "existing" to include subjects that exist, can plausibly exist or could plausibly have existed, and treat realistic AI-generated avatars and personas as covered. On that reading, a photorealistic avatar of an invented person falls within the definition.
Two caveats. Commission guidelines aren't binding law — only the courts can give a definitive interpretation of the Act. And the text of Article 3(60) itself says "existing", so until this is tested, treat a realistic invented avatar as a likely deep fake rather than a certain one. In practice the disclosure is cheap, so we'd label it.
A clearly stylised, cartoon or illustrated avatar is unlikely to "falsely appear to be authentic", so it generally won't be a deep fake — though if it chats with people, the 50(1) interaction disclosure still applies.
Where the avatar is part of an evidently artistic, creative, satirical or fictional work, the 50(4) disclosure can be limited to flagging that generated content is present, in a way that doesn't hamper the work.
| Setup | Likely rules | What it means |
|---|---|---|
| Photorealistic avatar that holds live conversations (support agent, video-call assistant) | 50(1) · 50(4) · 50(2) | Tell users they're talking to an AI, and disclose the avatar is artificially generated. The avatar vendor marks the output. |
| Cartoon or clearly stylised mascot that chats with visitors | 50(1) | Interaction disclosure still applies unless it's genuinely obvious it's AI. Unlikely to be a deep fake, as it wouldn't pass as authentic. |
| Pre-rendered AI presenter in a marketing or training video | 50(4) · 50(2) | No live interaction, so 50(1) doesn't apply — but a realistic synthetic presenter needs a deep fake disclosure. |
| Digital replica of a real person (your CEO, a spokesperson, a celebrity) | 50(4) · 50(2) | Squarely within the deep fake definition. Add 50(1) if the replica also converses with people. |
Whichever rules apply, visible disclosures must be clear and distinguishable and given at the latest at the first interaction or exposure — a watermark in the file isn't something a viewer can see. If your avatar is really a chat widget with a face, start with does my chatbot need an AI disclosure?
Usually, and often under more than one rule. An avatar that talks with people triggers the Article 50(1) interaction disclosure; the system generating its video or audio must mark that output under Article 50(2); and a realistic avatar that could pass as a real person can be a deep fake requiring disclosure under Article 50(4).
Probably, if it's photorealistic. Article 3(60) covers content that resembles existing persons and would falsely appear to be authentic. According to the widely reported interpretation of the Commission's final Article 50 guidelines (July 2026) — reported by multiple law firms including Bird & Bird, Reed Smith and Davis+Gilbert — "existing" includes people who could plausibly exist, so a photorealistic avatar of a person who doesn't exist can fall within the definition. The guidelines aren't binding, and this hasn't been tested in court. A clearly stylised or cartoon avatar is unlikely to qualify.
Only for Article 50(2), which is the vendor's obligation as provider. Machine-readable marking doesn't tell a visitor anything, so it doesn't replace the visible disclosures under Article 50(1) and 50(4).
Article50.io is an automated Article 50 transparency assessment platform that scans websites for potential EU AI Act transparency obligations and provides remediation guidance, implementation instructions, and compliance-ready disclosure language.
The free scan shows your single most severe finding in about 30 seconds — no signup, public pages only.
Automated technical guidance, not legal advice. Citations refer to Regulation (EU) 2024/1689.