The firebreak
An engineered barrier keeps the AI to information and navigation. It never diagnoses, interprets or triages. Anything urgent triggers immediate 999/111 signposting and hand-off to a human.
Trust & safety
Safety is engineered in — not promised
Every protection below is a mechanism, not a policy statement. They are designed in from the first line of code and reviewed by clinicians before anything reaches a real patient.
Four standing protections
NHS clinical safety standards
DCB0129/0160 from day one: a named Clinical Safety Officer, a hazard log, and a safety case signed before anything reaches a real patient.
Provable consent
Permissions are granted per purpose, withdrawable at any time, evidenced by standards-based receipts, and held on a tamper-evident audit trail.
Private by design
Processing comes to the data inside a secure boundary — the same protective model the NHS uses for its own secure data environments. Identifiable data is never sent elsewhere for AI processing.
What Wellness IQ is not
It is not a diagnosis tool. It does not replace GPs, nurses or any clinician. It is not a replacement for existing NHS record systems or shared care records — it works alongside them. It is not an emergency service, and tells its users so. And it is not a medical device under UK law: it is deliberately engineered to stay on the information side of that line, with a standing review against the MHRA's evolving AI framework.