Knowledge Base Regulation Tracker Standards & Frameworks OSS Directory AI Incident Database Research Paper Library Case Studies Society Articles News Jobs About Charter Sign inBecome a Member
Home/News/August 2026

EU AI Act transparency duties apply and the AI Office gains enforcement powers over general-purpose model providers

Providers must mark synthetic output in a machine-readable way, chatbots must disclose that they are AI, deployers must label deepfakes, and the AI Office can now investigate and fine GPAI providers, with a December deadline for systems already on the market.

Enforcement and litigationEuropean Union2 Aug 2026

Article 50 was untouched by the Digital Omnibus. The Code of Practice on marking and labelling finalised in June sets the expectation that metadata, watermarking and detection interfaces are combined, because metadata alone does not survive re-encoding. The AI Office's exclusive competence now also covers AI systems built on a GPAI model by the same provider.

What to do: implement provenance marking for generated media and text, add chatbot disclosures, audit your labelling against the code, and, if you supply a systemic-risk model, expect the AI Office to ask for the Model Report.