The Dutch Data Protection Authority (AP) details the new transparency obligations applicable to artificial intelligence systems from August 2, 2026.
From this date, users must be informed when they interact with an artificial intelligence system, such as a conversational agent, in order to distinguish it from a human interlocutor. Content generated or modified by AI, including images, videos, sounds, and texts, must be clearly identified as such, notably to recognize deepfakes. This identification may be done by technical means such as a digital watermark. Furthermore, any use of AI for emotion analysis or categorization of individuals based on their biometric characteristics must be indicated no later than at the first interaction.
According to the Algorithm Coordination Directorate (DCA) of the AP, these rules aim to enable individuals to better assess the information they receive, knowing for example if they are communicating with an AI on a dating app, if a video of a political figure has been modified, or if the voice responding in the name of a service is synthetic. The objective is to combat misinformation, identity theft, and fraud.