EU AI Act: labelling AI content is enforceable now — the 2 December marking deadline is next

The EU AI Act's transparency obligations for AI-generated content have been enforceable since 2 August 2026. Generative systems placed on the market before that date have until 2 December 2026 to add machine-readable marking, under the Commission guidelines adopted on 20 July 2026.

The rules sit in Article 50. Providers of generative AI systems must mark synthetic audio, image, video and text so detection tools can identify it as artificial. Deployers — the organisations using those systems — must disclose deepfakes and label AI-generated text published to inform the public.

What applies, and what does not

Two things get conflated. The Code of Practice on Transparency of AI-generated Content is voluntary; the Article 50 obligations are not. And the 2 December extension covers only systems already on the market before 2 August 2026 — anything launched after that date must comply from day one.

There is also a carve-out worth knowing: where a human takes editorial responsibility for a text, the deployer labelling duty does not arise. That is a real exception for written content, not for images, audio or video.

Do this week

If your team publishes webinars, courses or campaigns, inventory which assets contain AI-generated material, set your labelling rule, and ask each generative tool vendor one question: how do you implement Article 50 machine-readable marking, and by when?

Sources

• European Commission — https://digital-strategy.ec.europa.eu/en/policies/code-practice-ai-generated-content
• EU Artificial Intelligence Act — https://artificialintelligenceact.eu/implementation-timeline
• Stibbe — https://www.stibbe.com/publications-and-insights/the-ai-acts-transparency-obligations-rules-scope-and-timeline
• Orrick — https://www.orrick.com/en/Insights/2026/08/EU-AI-Act-Transparency-Obligations-for-AI-Generated-Content-Article-50