AI policy — assistance you can audit
We use AI where it reduces friction: drafting, summarizing, suggesting meals, accelerating logging. We do not use it to sneak medical autonomy past you, and we do not claim omniscience.
Principles
- Human agency first. Meaningful changes should be visible and reviewable by you.
- Plain-language limits. If a feature uses AI, the UI should say so without hunting.
- No “AI doctor.” Outputs are informational, not clinical decisions.
Data use for AI features
AI features process the inputs you provide and relevant saved settings to generate suggestions. We do not sell identifiable health journals to advertisers. We do not use your identifiable health logs to train public foundation models. If we ever offer opt-in research, it will be explicit.
When AI is wrong
Treat AI output as a draft. If a meal suggestion ignores an allergy, discard it. If a summary misreads your week, correct it. If stakes are high — pregnancy, medications, disease management — involve a qualified professional.