News / Technology

Tier 1 · Decisive

The EU AI Act: transparency rules took effect on 2 August

EU · Denmark

  1. Debated
  2. Proposed
  3. Adopted
  4. In force

Last reviewed · Last changed

Researched and written with AI · Method

The European Commission's Berlaymont building in Brussels, where the Commission's AI Office gained enforcement powers from 2 August 2026. Photo: Stern / Wikimedia Commons, CC BY-SA 3.0 · source

On 2 August 2026 the EU's transparency rules for artificial intelligence took effect, and the Commission's AI Office and the national authorities began enforcing the AI Act. Chatbots must tell the user they are an AI system, and certain deepfakes and other AI-generated content must be labelled, including in machine-readable form. The rules for high-risk systems have been postponed.

Decided by
The European Parliament and the Council; the Commission's AI Office and the national supervisory authorities enforce the rules.
Kind
Decision or incident
Reach in Denmark
Almost everyone
Applies from
2 Aug 2026
European Commission, Brussels · Open in Google Maps

© OpenStreetMap contributors (ODbL)

What has been decided

The AI Act, Regulation (EU) 2024/1689, entered into force on 1 August 2024 and applies in stages. Prohibited practices and AI literacy obligations applied from 2 February 2025. The rules on governance and on general-purpose AI models applied from 2 August 2025. From 2 August 2026 the transparency rules apply, and enforcement has begun.

The Commission wrote on 2 August 2026 that certain images, audio and video that resemble existing persons, objects, places or events (deepfakes) must be clearly labelled and carry machine-readable marks. The same applies to emotion-recognition and biometric-categorisation tools, and to text published to inform the public on matters of public interest where there has been no human editorial control. Users must be clearly informed when they are not interacting with a person but with an AI system, for example a chatbot, an AI agent or an avatar.

Fines can, according to the Commission, run up to 15 million euro or 3 percent of global annual turnover for companies, with proportionality for small and medium-sized enterprises, and up to 750,000 euro for EU institutions.

The rules for high-risk systems were amended by Regulation (EU) 2026/1744, the Digital Omnibus on AI. According to the Commission's timeline, the obligations for high-risk systems in certain sensitive areas, including biometrics, critical infrastructure, education, employment, migration, asylum and border control, apply from 2 December 2027, and the obligations for AI embedded in regulated products, such as lifts or toys, from 2 August 2028.

The Commission writes that the aim is to foster trustworthy AI, to help people recognise when they are encountering AI, and to reduce deception, while creating a single market and protecting health, safety and fundamental rights.

What it means for Denmark

The regulation applies directly in Denmark. Danish firms and public authorities that provide or deploy AI systems must meet the transparency rules from 2 August 2026. National market surveillance authorities enforce the rules together with the European AI Office. The Folketing is considering a proposal for a Danish AI act with supplementary rules on supervision and sanctions (L 3); the proposal has not been adopted. In the 2027 finance bill proposal the government at the same time sets aside an AI acceleration fund of 200 million kroner a year in 2027 and 2028.

What to watch

Will the Folketing adopt the Danish AI act, and which authorities will get the supervisory role? How will labelling work in practice for ordinary users? And will the postponed dates for high-risk systems, 2 December 2027 and 2 August 2028, hold, or will the timetable change again?

Sources

Changelog

  1. Headline clarified to name the subject
  2. Item created because the AI Act's transparency rules and enforcement took effect on 2 August 2026.