Free resource · Article 50
Disclosure templates you can paste in
The EU AI Act’s Article 50 transparency rules (live 2 August 2026) mostly come down to telling people, clearly, when they are dealing with AI. Here are plain-language starting points for the four most common cases. Not sure which apply to you? Run the transparency check.
AI chatbot / assistant
Article 50(1)Shown at the start of any chat where a user might think they are talking to a person.
You’re chatting with an AI assistant. It can make mistakes; for anything important, you can ask to speak with a member of our team.
AI-generated text
Article 50(2)On articles, summaries, or descriptions produced with generative AI.
Parts of this content were produced with the help of AI and reviewed by our team.
AI-generated image
Article 50(2)On or beside any image created or materially edited with AI.
This image was generated or edited using AI.
Deepfake / synthetic media
Article 50(4)On audio or video that has been AI-generated or synthetically altered to resemble real people or events.
This video contains AI-generated or synthetically altered content.
Emotion or biometric analysis
Article 50(3)Wherever an AI system infers emotion or categorises people by biometric data. Replace the brackets.
This service uses AI to analyse [facial expressions / tone of voice] in order to [purpose]. You can object or opt out at [contact or link].
Starting points, not lawyer-approved copy. Adapt the wording, placement, and opt-out mechanics to your product and get them reviewed. Emotion inference in the workplace or education, and certain biometric uses, are prohibited under Article 5 regardless of any notice. See our disclaimer.