Do you need to disclose AI-generated video in the European Union?

EU AI Act Art. 50 — transparency for AI-generated & deepfake content

Article 50 of the EU AI Act sets transparency duties: providers must mark AI outputs in a machine-readable way; deployers of deepfakes must clearly disclose that the content is artificially generated/manipulated; AI-generated text published to inform the public on matters of public interest must be disclosed; and users must be told when they interact with an AI system (chatbots). Transparency obligations apply from 2 Aug 2026.

Source: EU AI Act — Article 50 (transparency obligations) · reviewed 2026-06-09

YouTube requires disclosing realistic altered or synthetic content

When you upload content that is realistic and was meaningfully altered or generated with AI, you must tick 'Altered or synthetic content' in YouTube Studio. YouTube then shows a label in the description (and a more prominent label on the video for sensitive topics like health, news, elections, finance). Clearly unrealistic, animated, or trivially-edited content is exempt.

TikTok requires labeling realistic AI-generated content

TikTok requires creators to turn on the 'AI-generated content' toggle for realistic AI content. TikTok also auto-labels content that carries C2PA Content Credentials and content made with TikTok AI tools. Undisclosed realistic AI content may be removed under the Synthetic and Manipulated Media policy.

Meta labels AI content and expects creators to disclose

Meta (Instagram & Facebook) applies an 'AI info' label to content it detects as AI-generated or that carries industry signals (C2PA, IPTC). Creators are expected to disclose photorealistic AI content; failing to disclose materially deceptive AI media can lead to penalties.

X may label or limit deceptive synthetic & manipulated media

X's Synthetic and Manipulated Media policy targets media that is deceptively altered/fabricated and likely to cause harm. X may add labels, reduce visibility, or remove it; Community Notes often add context. There is no universal upload toggle, so proactive in-post disclosure is the safe practice.