How Kinetta uses AI

Draft (under review). Version 0.3-draft.

Kinetta uses AI in a small number of clearly marked places. This page explains where it appears, what it receives, and what it cannot do.

Where AI appears

Ask Kinetta: the only assistant in the app. It answers questions about your body, movement and recovery.
Discover: AI-written explanations alongside curated content and publicly described experiences.
What Kinetta remembers: suggested summaries of things you have told the app, which you confirm before they are kept.

Every AI-generated output carries a visible AI-generated indicator.

What AI receives

The question or content you are working with, and relevant context from your record where you have allowed it. It does not receive more than is needed to answer.

What AI never receives

Patient Voice is excluded from AI

Patient Voice is a private channel between you and your clinician. Nothing written there is read, summarised or processed by AI, including in aggregate. This is enforced in the data layer, not just hidden in the interface.

What AI produces, and its limits

Educational information, not personalised medical advice.
It can be wrong, incomplete or out of date.
It does not diagnose and does not prescribe treatment.
It is not an emergency service and is not monitored in real time.
Publicly described experiences are anecdotal. They are shown separately from Kinetta's own explanation and are never presented as evidence or recommendation.

Reporting an unsafe or incorrect answer

Every AI answer has a report action. Reports are reviewed as part of AI safety monitoring. You can also contact [SUPPORT EMAIL]. If something is urgent, do not wait for a response; contact a qualified health professional or emergency services.

Provider, processing and retention

Kinetta is designed to be model-agnostic. The provider can change, and the app does not depend on a single vendor.

AI provider: [AI PROVIDER]
Processing location: [AI PROCESSING LOCATION]
Provider retention setting: [RETENTION SETTING]
Provider model-training setting: [MODEL TRAINING SETTING]
Subprocessors: [SUBPROCESSOR LIST]

These details will be updated when Kinetta’s production providers and contractual settings are confirmed. Kinetta will not state that a provider does not retain or train on information until the provider, contract, plan and settings have been confirmed.

Human responsibility

Where you are in care, your practitioner remains responsible for clinical decisions and clinical records. AI does not make clinical decisions in Kinetta.

Read the Privacy Policy for how information is handled more generally.