Where artificial intelligence is involved in this service, what it does, what it does not do, and how to report a problem.
In effect since
1. Where AI is involved
In two places, both explicit:
- Guided sessions. A conversational agent runs the session by voice, following the Méthode Dolfino’s scripts.
- The tutor. A written assistant that answers questions about course content.
Everywhere else — the site’s pages, the courses, the articles — the content is written by the school.
2. How this works
A session is conducted by a third-party language model, governed by the method’s scripts and rules. The audio flows in real time: it is transcribed to be understood, the reply is spoken, and the audio is not retained.
The transcript is kept only if you have given the corresponding consent, and it disappears if you withdraw it.
3. What the AI does not do
- It makes no diagnosis and gives no medical advice.
- It makes no automated decision producing a legal effect or significantly affecting you. It does not grade you, does not classify you, and does not condition any access.
- It does not replace a teacher. Certifying trainings are given by certified teachers.
- It does not learn from you. Your exchanges are not used to train our providers’ models.
4. The safeguards
Screening. A prior questionnaire conditions access to AI sessions. It exists because a visualization session is not suited for everyone, at all times.
Closed by default. If the screening cannot be verified, no session starts. If consent to the live voice is not given, no session starts.
Limited scope. Some techniques in the method are taught but never conducted in an AI session; the course pages indicate this scope.
5. Reporting problematic behavior
If the agent behaves badly — an inappropriate remark, advice it should not have given, insistence after a refusal — write to reussite@dolfino.tv stating the approximate date and time of the session.
You are the only possible witness to this kind of misbehavior. These reports are handled as a priority.
