AI use policy
Version in force on 21/11/2026
1. Our commitment
OceanaCode designs AI employees: conversational assistants based on artificial intelligence that answer an agency's customers on its behalf. This policy explains in plain words how they work, what they do with your data and the limits we set.
2. Transparency: you know you are talking to an AI
Every AI employee clearly states, below the conversation, that you are chatting with an assistant based on artificial intelligence, in line with Article 50 of Regulation (EU) 2024/1689 on artificial intelligence. It never pretends to be a person.
3. A human can always take over
The agency's team sees the conversations and can take them over at any time. The assistant hands over to an advisor when you ask, in case of a complaint or when it is unsure of its answer. Appointments it offers are confirmed by an advisor.
4. What the assistant knows, and what it does not do
The assistant answers only from the information provided by the agency (properties, fees, opening hours, frequent questions). It gives no legal, tax or financial advice, promises neither prices nor availability, and makes no decision with legal effects on you. Its answers are generated automatically and may contain errors: the agency remains responsible for its commitments.
The assistant does not treat people differently based on their origin, nationality, family situation or any other discriminatory criterion, and refuses such requests.
5. Your data
Messages are processed to answer your request and pass your project on to the agency, which is the data controller; OceanaCode acts on its behalf as a processor. To produce each answer, the message is sent to the AI model provider (currently Cloudflare Workers AI) under its contractual terms. Your messages are not used to train artificial intelligence models.
Data is hosted in France (OVHcloud) and kept for as long as needed to follow up your request. You can exercise your rights (access, rectification, erasure, objection) with the agency or by writing to contact@oceanacode.com. Details are in the privacy policy.
6. Quality and security
Before each launch and every change, the assistant is tested on dozens of simulated conversations, including attempts to misuse it. Conversations are monitored to correct imperfect answers. If the model is unavailable, a fallback message invites the person to be contacted by the agency.
7. Questions and reporting
For any question about this policy or to report an inappropriate answer: contact@oceanacode.com. This policy may change with our services and the regulations; the version date appears at the top of the page.