AI Transparency
What the avatar really is, how it works and what its limits are.
Last updated: 29 May 2026
Transparency notice under Art. 50 of the EU Artificial Intelligence Regulation (AI Act).
1. You are interacting with an AI
The avatar is a generative artificial intelligence system. Every response is produced automatically by a language model. It is not the real person and does not communicate with you on their behalf.
2. How it works
When you create an avatar, the memories you upload are split into fragments and turned into numerical representations (embeddings). During a conversation, the system retrieves the fragments most relevant to your question and provides them to the language model as context, so that the response stays anchored to the real materials rather than being entirely invented.
Responses are generated by third-party providers' models (see the privacy policy for the list).
3. Limits and possible errors
- The avatar may generate inaccurate, incomplete or entirely fabricated statements ("hallucinations").
- It does not represent the real thoughts, opinions or will of the person.
- Quality depends on the uploaded materials: sparse memories produce more generic responses.
- Responses should not be used for legal, medical, financial or other significant decisions.
4. Emotional care
This tool touches an intimate and painful experience like grief. Use it in the time and way that feels right for you. It is not a substitute for the support of loved ones or professionals; if you feel the need, seek qualified help.
5. Generated content
Texts, letters and audio content produced by the avatar are artificially generated. We ask you not to present them to others as authentic statements of the person represented.