Privacy Policy
Last updated: 25 July 2026
audioFirst is operated by Agonist Development AB (Sweden). This policy explains how the audioFirst website and iOS app handle personal information. The app is privacy-first: study progress and card content stay on your device by default.
1. What we collect
- Website email — if you contact us or join a waitlist, we store your email address to reply or notify you about launch.
- Anonymous app usage analytics — the iOS app may send product interaction events such as session started, session completed, card graded, pack opened, paywall shown, or settings export. We do not send card content, audio recordings, transcripts, your name, phone, location, or device serial.
- Subscription state — if you subscribe, Apple processes payment through StoreKit and Adapty helps us manage entitlement state. Adapty receives StoreKit receipt data, an anonymous user ID, subscription tier, expiration date, and renewal status.
- Account (Sign in with Apple) — optional. Bundled and personal packs work without an account. If you choose to sign in (Settings → Account), the app requests the name and email you agree to share via Sign in with Apple (you can share your real email or an Apple private-relay address). Our backend verifies Apple's identity token to issue a session token and determine which private course packs your account can access. This exchange happens once per sign-in or token refresh.
- Live Voice Tutor audio — optional and session-scoped. If you explicitly start a Live Voice Tutor session on a pack, your spoken audio is streamed from your device to our backend and relayed to a third-party realtime AI voice provider for the duration of that session only, so it can converse with you and grade your spoken answers. This is a separate, opt-in mode — the default flashcard and hands-free study modes never send audio anywhere. See §4 and §5.
- Local study data — study progress, session history, and preferences are stored on your device. If you uninstall the app, iOS deletes this data.
2. Why we use it
We use this data to operate audioFirst, reply to support requests, provide subscriptions, improve the product, and meet legal obligations. We do not sell personal data or use it for cross-app tracking.
3. Legal basis (GDPR)
We process data under your consent, the performance of our contract with you, our legitimate interest in improving the product, and legal obligations where applicable.
4. Device permissions
- Microphone — used for two things. (1) During a hands-free study session, to grade cards with voice commands such as "good", "hard", or "again" — this stays on your device and audio is not stored by us. (2) If you explicitly start an optional Live Voice Tutor session, your spoken audio is streamed to our backend and relayed to a third-party realtime AI voice provider for the duration of that session only (see §5). Live Voice Tutor is a separate opt-in mode; ordinary study never sends audio off your device.
- Speech Recognition — required by iOS for Apple's speech-recognition framework, used to understand voice commands. Speech recognition runs on-device when possible; Apple's privacy policy governs anything Apple processes. (During an opt-in Live Voice Tutor session, the AI voice provider does its own speech understanding server-side — see §5.)
- Motion & Fitness — optional. If you enable the Learning Context feature, the app reads iOS motion-activity data (whether you are stationary, walking, running, or in a vehicle) to compare how well you recall in each context. This is processed on your device, is never sent to us, and the feature is off until you turn it on.
- Photos — optional. If you allow it, the app watches for new screenshots so it can offer to turn the text in them into flashcards. Image text recognition runs on your device, your photos are never uploaded to us, and you can disable this in Settings.
5. Sharing
We use trusted infrastructure and service providers only to run the product:
- Cloudflare — website hosting and routing.
- Apple — App Store distribution, payments (StoreKit), speech-recognition services, and, if you choose it, Sign in with Apple (name/email you agree to share).
- Adapty — subscription entitlement management.
- Realtime AI voice provider — only if you start an opt-in Live Voice Tutor session. Depending on availability the provider is one of Microsoft Azure OpenAI Service, OpenAI, or Google Gemini (realtime voice APIs, cascade order Azure → OpenAI → Gemini, Azure first for EU data residency). It receives your streamed spoken audio and the current card text for the duration of that session only. Provider API keys are held only on our backend, never on your device.
- Self-hosted PostHog — anonymous product analytics, run on our own server (no third-party analytics SaaS).
The shipped app does not call any AI API at runtime for ordinary study — flashcards, hands-free voice commands, screenshot text recognition, text-to-speech, and translation all run on-device or against our own non-AI backend. The one exception is the optional, explicitly-started Live Voice Tutor session described above.
6. Your rights
You can request access, correction, deletion, or export of your data, and withdraw consent at any time. Email privacy@audiofirst.org.
In the iOS app, you can export your local data as JSON or delete all local study data from Settings.
7. Retention
We keep support and waitlist emails as long as needed to respond or notify you, unless you ask us to delete them. Local study data stays on your device until you delete it or uninstall the app.
8. Contact
Data controller: Agonist Development AB. Privacy contact: privacy@audiofirst.org.