EU AI Act Article 50 is now enforceable — chatbots, deepfakes and machine-readable marking
Article 50 transparency obligations became generally applicable across the EU on 2 August, enforceable by national authorities, with the Commission's AI Office simultaneously gaining investigation and enforcement powers over general-purpose model providers. Users must be told when they are talking to an AI, synthetic output must be machine-readable and detectable, and deepfakes must be labelled. Penalties reach €15 million or 3% of worldwide turnover.
The machine-readable marking requirement is the one with real engineering weight. Telling a user they are speaking to a chatbot is a disclosure string. Making every generated image, video, audio clip and text output detectable as synthetic in a machine-readable format is a provenance pipeline, and it has to survive re-encoding, cropping and platform processing to be worth anything.
Providers with generative systems already on the market have until 2 December to comply with marking and detection specifically. Everything else in Article 50 is live now. That staggering is a tell — the regulator understood that retrofitting provenance into a shipped product is a different order of task from adding a disclosure banner.
The AI Office's new enforcement powers over general-purpose model providers are the structural change underneath the headline. Article 50 sets duties; the enforcement authority determines whether they are real. Watch the first investigation rather than the first compliance blog post — and note it lands the same week the US chose not to publish its framework at all.
European Commission — Commission starts enforcing AI Act rules and new transparency requirements on 2 August → · Cooley — EU AI Act: transparency obligations take effect 2 August 2026 → · Goodwin — Not delayed, not deferred: EU AI Act transparency obligations are now in force →