Data & AI use
Document version: 2026-09-21-draft-1
Preview document — not yet in effect. Registration remains closed while operator details and legal review are completed.
An optional assistant
The public AI chat provides general information and helps you prepare questions. It does not diagnose, prescribe, verify clinicians or retrieve current ratings. Provider comparisons and case analysis are planned capabilities, not claims about the current chat.
What is sent
After you choose to use the chat, your submitted messages and a limited recent conversation history are sent through Adamiani to OpenAI to generate an answer. The interface language and coarse country context may also be included. Your account profile is not automatically attached.
Your choice
The chat asks separately for agreement before sending messages. You can leave this choice unchecked and still create an account. Unchecking it stops future chat requests; it does not recall information already processed. Clearing the on-screen chat removes its visible history, not any service-provider logs.
Keep sensitive information out
Do not send your full name, contact details, identification numbers, medical documents or identifiable case histories. Use general questions. AI can make mistakes; check important information with a qualified clinician and use local emergency services when necessary.
Storage and service-provider handling
The current integration requests OpenAI responses with store:false and does not save chat history in the account database. This does not mean zero retention by infrastructure or AI providers. OpenAI's API data controls explain provider handling. The final retention and transfer arrangements must be reviewed before sensitive clinical features are introduced.