Q&A
Transparency requirements for AI: what does this mean for you?
Public explanation by the Dutch Data Protection Authority of the transparency rules that have applied since 2 August 2026, with a downloadable infographic. Four situations: AI systems such as chatbots must be recognisable, AI-generated content receives a digital marking, people are informed about emotion recognition and biometric categorisation, and deepfakes and unedited AI articles on matters of public concern receive a visible label. The Authority confirms that systems placed on the market before 2 August 2026 have four additional months for the marking obligation, namely until 2 December 2026. Living topic page, updated on 31 July 2026.
What this means for you
This is the Dutch supervisor's explanation of the transparency obligation that already applies, including the additional reprieve for existing systems.