Privacy Policy
Last updated: February 2026
Data We Collect
We collect minimal data to make Swift English Vocabulary work. Here's exactly what we store:
- Email address — for account recovery and authentication. If you sign in with Apple, your email and name are provided by Apple for account creation. Apple may provide a private relay email address.
- Display name — to personalize your experience
- Learning data — words you've studied, review statistics, study streaks, and achievement progress
- Child profiles — name and avatar for parent-managed multi-profile accounts
- Optional PIN hash — if you enable PIN protection for child profiles (stored securely, not readable by us)
- APNs device token — if you enable push notifications, your Apple Push Notification service token is collected and shared with Crisp solely for delivering support chat notifications
- FCM device token — stored in Firestore, used by Firebase Cloud Messaging for push notifications (updates, reminders). One token per account.
- Feature requests — title, description, and display name visible to all users. Votes recorded with your user ID.
- AI Tutor conversations — if you give explicit consent, chat messages with the AI Tutor are stored in Firestore under your account. Conversations include your messages, AI responses, and contextual learning data. The following data is shared with Anthropic to generate tutor responses: chat messages and conversation history, words being studied (word, definition, example, phonetics), learning statistics (total words, mastered words, streak), and your translation language preference. No data is shared until you explicitly agree.
- AI data sharing consent — your consent status and the date you gave consent are stored in your user profile
Third-Party Services
Firebase
We use Firebase for authentication and as our database backend (Firestore). Firebase is a Google service that stores your account data and learning progress.
Firebase Crashlytics
We use Firebase Crashlytics to collect crash reports and diagnostic data to improve app stability. Crashlytics automatically collects device information (model, OS version), app version, and the screen you were on when a crash occurred. It also logs non-fatal errors with diagnostic context (subscription status, current screen name). Your Firebase user ID is linked to crash reports for correlation. No personal content (words, progress, or learning data) is included in reports. See Google's privacy policy.
Firebase Cloud Messaging
Delivers push notifications. Receives your FCM device token. Your device is subscribed to topic groups (subscription tier, app version, activity status) for targeting. No personal data beyond the token is sent.
Firebase Remote Config
Fetches app configuration (e.g. free tier word limits) from Firebase. No personal data is sent; Firebase sees the fetch request.
RevenueCat
RevenueCat manages subscriptions. We share only your App Store receipt information for subscription processing and validation.
Crisp
Crisp provides in-app support chat. When you use the support chat, Crisp receives your email address, display name, and APNs device token (for push notifications about support conversations). If you share images in a support conversation, Crisp accesses your camera or photo library for that purpose. See Crisp's privacy policy.
Free Dictionary API
We fetch word definitions from the Free Dictionary API. When you look up a word, we send that word to their service to retrieve pronunciation, definitions, and examples.
Anthropic
The AI Tutor is powered by Anthropic's Claude language model. Data is only shared with Anthropic after you give explicit consent through an in-app disclosure. When you use the AI Tutor, the following data is sent to Anthropic's API: your chat messages and conversation history, words being studied (word, definition, example, phonetics), learning statistics (total words, mastered words, streak), and your translation language preference. Anthropic does not use this data for training. See Anthropic's privacy policy.
What We Don't Collect
Swift English Vocabulary is designed to respect your privacy. We explicitly do not:
- Track behavior — we don't track user behavior, session duration, or feature usage for marketing or analytics purposes. We do collect crash reports and basic diagnostic data via Crashlytics for app stability.
- Show ads — Swift English Vocabulary is ad-free
- Track location — we never request or store GPS data
- Device tracking — we don't collect device identifiers for tracking or advertising. Crashlytics collects device model and OS version for crash diagnostics only. An APNs device token is shared with Crisp for support chat push notifications; an FCM device token is used for app push notifications via Firebase Cloud Messaging.
Your Choices
You have control over how your data is used:
- AI Tutor consent — before any data is shared with Anthropic, you must explicitly agree through an in-app data sharing disclosure. The AI Tutor is not accessible until you consent.
- Revoke consent — you can revoke your AI data sharing consent at any time from your profile settings. Revoking consent permanently deletes your entire chat history and disables access to the AI Tutor.
- Account deletion — you can delete your account at any time from the app settings, which permanently removes all your data
Translations
Swift English Vocabulary uses Apple's on-device translation framework to provide word translations. This means:
- Translation happens entirely on your device using Apple's built-in models
- No translation data is sent to external servers or third parties
- Your words and translations remain private to your device
Data Storage
Your account and learning data are stored on Firebase, which uses Google Cloud infrastructure. Firebase provides enterprise-grade security including encryption in transit and at rest.
Data is backed by Google's data centers with automatic redundancy and disaster recovery.
Account Deletion
You can delete your account at any time directly from the app settings. When you do:
- Your entire account and all data are permanently deleted
- All child profiles linked to your account are also deleted
- Deletion is immediate and cannot be undone
Children's Data
Swift English Vocabulary supports multi-profile accounts so parents can manage learning for their children. Here's how it works:
- Child profiles are created and managed entirely by a parent account
- Child profiles store only a name and avatar (no email or other personal information)
- Swift English Vocabulary does not directly collect data from children or require consent from children
- Crash diagnostics (crash reports, device model, OS version) are automatically collected for all profiles including children's, but contain no personal data beyond what is already disclosed above
- All data is stored under the parent's account and is managed by the parent
- Feature request submissions are not available for child profiles. FCM token and push notification topics apply at the account level, not per child profile.
- AI Tutor conversations from child profiles are sent to Anthropic's API in the same manner as adult profiles. The parent's consent (given via the parent account) applies to all child profiles under that account.
Questions or Concerns?
For general support, reach out via the in-app Crisp chat. For formal or legal inquiries, contact us at:
The Bottom Line
Swift English Vocabulary respects your privacy. We collect minimal data, don't track you, and don't sell your information. Your learning is yours alone. We believe language learning should be a personal journey, not a data collection opportunity.