← cleanmy.life

COPPA Disclosure

CleanMy COPPA Disclosure

Effective date: to be set at publication. Published at: https://cleanmy.life/coppa (or as a section of the main privacy policy).

This document is the source for our COPPA disclosure under 16 CFR §312. It can live at its own URL or as a section of the main privacy policy depending on what we decide is more discoverable. Apple's reviewer cares mostly that it exists, that it explains parental consent, and that the parent can easily revoke.

The shorter consumer-facing version is in privacy-policy.md "Children's privacy" section.


COPPA Disclosure for CleanMy

CleanMy is used by both adults and children. This disclosure describes the practices required by the Children's Online Privacy Protection Act (COPPA, 15 U.S.C. §§6501-6506, 16 CFR Part 312) and similar laws (GDPR-K in the EU, state laws in California, Connecticut, Maryland, New York, Texas, etc.).

Who this applies to

This disclosure applies if your child under 13 is using CleanMy in the family configuration. In the family configuration:

If an under-13 child set CleanMy up themselves without a parent doing the in-person setup, that is not the family configuration, and we do not authorize that use.

What information we collect from children

In the family configuration, we collect from the child's device:

TypeWhyWhere storedRetention
Cleaning photos (before/after)AI task generation + verificationSupabase Storage, linked to the kid's anonymous user UUIDUntil parent or kid deletes
Anonymous user UUIDApp functionalitySupabase user tableUntil parent or kid deletes
Cleaning session metadata (task list, completion timestamp, AI judgment)Bank of earned screen-time minutesSupabase tablesUntil parent or kid deletes

We do NOT collect from the child:

The child-facing UI contains no fields where the child could enter such information. We never ask a child to type their own name into the app.

Child's first name — provided by the parent

The setup flow asks the parent to type the child's first name. This name is:

When the parent reviews and publishes one of the child's cleans to the public feed, our server creates a separate public post owned by the parent. The public post's displayed label is the parent's chosen public name; if the parent has not set one, the post appears under a system-assigned user handle (for example, user_a4b3c2d1) derived from a one-way hash of the parent's account ID. The child's first name is never used as the public post label. If the parent chooses to mention the child by name in the free-text caption they write themselves at publish time, that is the parent's own speech and is outside what we collect.

We treat a first name standing alone as something other than the per-child personal information COPPA's Rule (16 CFR §312.2) is built around. The Rule's enumerated list of "personal information" requires "first and last name" for the name field, and standalone first names are not in the list. We do not store any child's last name. Because the first name is never published alongside a public persistent identifier (the public post uses the parent's chosen name or a system-assigned user handle), the §312.2(11) combination test does not attach. The internal pending-review record IS server-side, but it is parent-scoped, never public, and exists for the sole purpose of letting the paired parent identify which child's clean is in front of them when they review.

A note on the Quick Clean timer: when a child uses the Quick Clean fallback (a timer-based cleaning session with no AI verification), the only data we record about that session is an anonymous bank-event timestamp and the minutes earned. No photo is captured. No AI call is made. No content leaves the device. A child who only ever uses Quick Clean has shared essentially nothing about themselves with our server.

How we obtain verifiable parental consent (16 CFR §312.5(b))

The COPPA Rule's §312.5(b)(2) "reasonable efforts" standard requires the operator to use a consent mechanism "reasonably calculated, in light of available technology, to ensure that the person providing consent is the child's parent." CleanMy's mechanism, end to end:

  1. Adult Sign in with Apple. To set up the family configuration, the parent installs CleanMy on the parent's own device and authenticates via Apple's Sign in with Apple. We hold a server-side record of this verified adult identity, along with the parent's email (real or Apple relay address).
  2. Parent generates a pairing code on their device. CleanMy on the parent's device produces a single-use 6-digit pairing code that the parent then types into CleanMy on the child's device. Only an adult with control of the parent's SiwA-verified device can produce this code; the kid device redeeming it is then permanently associated with the parent's account in our database.
  3. Per-publish review and approval. A child's clean can never publish on its own. A database rule forces every child-captured post into a pending-review state, regardless of what the rest of the system requests. For every individual instance in which the parent chooses to publish one of the child's cleans to the public feed, the parent must review and approve it on the parent's own Sign in with Apple verified session. Our server records the approving parent's verified identity and makes that parent the publisher of record for the post, providing an auditable per-act consent record.

CleanMy does not at this time make programmatic use of Apple's Family Sharing APIs to independently verify family relationships, and we do not hold or process the parent's payment card. The consent mechanism above is what we rely on; it is what the FTC's "reasonable efforts in light of available technology" standard contemplates, and the per-act parent review and approval on the parent's verified session goes meaningfully beyond a one-time-consent posture.

How parents can review, delete, or refuse to consent to further collection

A parent can, at any time:

We do not use children's information for advertising

We do not serve ads to a child using CleanMy. Even if we later add an ad-supported free tier on the adult product, children's data is never used to serve ads, and ads are never shown to a child using the kid surface of CleanMy. We know which users are children because the kid surface is a distinct mode of the app, set during pairing by a Sign-in-with-Apple-verified parent and stored tamper-resistantly on the device. We do not share children's information with any advertising network. We do not use any cross-app tracking or device-identifier collection on the kid surface.

We do not enable communication between children and other users

The child-facing surface in CleanMy contains:

Children using CleanMy cannot communicate with strangers through the app.

Third-party service providers that touch children's data

ProviderWhat they receiveWhyWhere
Supabase Inc.Encrypted photos + UUID + session metadataBackend hostingUS
Google LLC (Vertex AI service)Photo content for AI task generation + verification. We use Gemini 3.5 Flash today; Vertex AI's catalog includes other multimodal models (Google Gemini 3.1 Pro / Flash-Lite / Pro Image, Anthropic Claude 4.x, Mistral Small 3.1 + OCR 2505) we may evaluate. Vertex AI's terms commit Google to not using customer content to train Google's models or any partner-model provider's models. AI consent for kid-captured photos is inherited from the SiwA-verified parent; a kid never grants their own AI processing consent.AI vendorUS
Industry image-hash database providers (e.g. Microsoft PhotoDNA, planned for V1.5)Perceptual photo hashes only, never photo contentModeration hash matchUS (planned)
SentryCrash diagnostics (no PII, no photo content)DiagnosticsUS
PostHogAnonymous product-interaction events (no PII, no photo content)AnalyticsUS

We require each provider, where contracts apply, to use the data only for the purpose above and to delete it when we terminate the relationship. Where the provider's published terms govern (Google's Vertex AI service in particular), we describe those terms accurately above rather than overstating the protections in place.

State-law overlays

Contact

For COPPA-related inquiries:

We respond to COPPA inquiries within 7 days.