Privacy Policy

Guardian Angel is an AI safety companion available to people of every age. What we collect, how we protect it, and who can see it depends on the type of account you hold. This policy describes those differences plainly.

Account Types

• Parent/guardian accounts manage one or more minor profiles (ages 5–17).
• Individual accounts are held by adults 18+ for their own use (young adult, adult, or older adult).
• Caregiver accounts are invited by an older adult and have view-only access to a limited safety signal view.

The Diary Model (parent ↔ minor)

When a parent manages a minor's profile, the minor's conversation content is private. Parents see safety signals, risk levels, detected patterns, and engagement metadata — but they cannot read what the minor writes. The only exception is an active, serious danger event, where specific situational details may be escalated to the parent and to the crisis resources the product routes to.

This model preserves the trust that makes the tool effective for minors. It does not apply to individual adult accounts, which do not have a parent dashboard.

Caregiver View (older adult ↔ caregiver)

An older adult can invite up to three caregivers. Caregivers see a narrow view: critical safety alerts (such as a suspected scam the older adult explicitly chose to share) and an engagement summary. Caregivers do not read conversation content. The older adult can revoke a caregiver at any time from their settings.

What We Collect

• Account information: name, email, and (for adult accounts) age.
• Minor profile information (name, age, avatar) created by a parent.
• Conversation content between the user and Guardian Angel, encrypted at rest.
• Safety metadata: risk levels, detected patterns, escalation events.
• Support-network or trusted-adult contact names and relationships that the user provides.
• Stripe billing identifiers for paid accounts. We do not store card numbers.

How We Use Data

• To provide real-time safety evaluation of conversations.
• To surface risk levels and patterns to authorized reviewers (parent for a minor's account; caregiver where explicitly granted).
• To route users to crisis resources when active danger is detected.
• To improve safety classification accuracy. Aggregate, de-identified data may be used for this purpose; conversation content is never sold.

Data Retention

Retention is user-configurable (30, 90, 180, or 365 days). On parent-managed accounts, the parent sets retention. On individual adult accounts, the user sets retention. Data older than the chosen window is automatically deleted. You can export or delete your data at any time through Settings.

Data Security

Data is stored in Supabase with row-level security. Parents can only access their own children's profiles; caregivers only see what the older adult has granted; individual users can only access their own data. Conversation content is blocked from direct parent and caregiver access at the database level.

Your Rights

You can request access, correction, export, or deletion of your data at any time. Account deletion removes all associated data, including conversations and safety events. Residents of California, the EU, and other jurisdictions with additional privacy rights can exercise them by emailing the address below.

Children Under 13

Minor profiles under 13 are created only by a verified parent or legal guardian, in line with COPPA requirements. A child under 13 cannot create their own account.

Contact

Questions about privacy? Email us at privacy@guardianangel.app