Security & Data Protection

Lunas uses authenticated API routes, server-side plan and ownership checks, signed file access, content validation, audit logs, and rate limits for sensitive actions.

The product is designed to minimize structured PII in health records where possible. Account data, clinical free text, timeline notes, and uploaded files are treated as sensitive data.

Files are stored outside the app database and served through ownership-checked access paths. Production storage uses private object storage with short-lived signed access.

Deleted, quarantined, or unauthorized records are excluded from reads, search, RAG retrieval, and sharing surfaces.

Security-sensitive actions are designed to be audit logged with IDs, statuses, and versions, not prompts, file text, model outputs, or medical values.

Training data requires consent, de-identification, and human approval before use in a model-improvement dataset.

Lunas runs on HIPAA-ready infrastructure: business associate agreements are executed with Google Cloud, Neon, and Vercel — every vendor in the health-data path. HIPAA readiness describes our infrastructure and contracts, not a certification; formal compliance status is validated before any regulated claim is marketed. SOC 2-style controls, retention rules, and clear user rights continue to mature.

Direct-to-consumer personal health record features and regulated Circle/Lab deployments can trigger different legal obligations, including breach-notification rules. Lunas reviews these obligations before expanding regulated use.

Report security concerns to support@lunas.one.