FameSim iconFameSim

Privacy Policy

Effective: April 21, 2026 · Last updated: April 21, 2026

FameSim ("we", "us") is an iOS live-stream simulator published by Appnor. This policy describes what data we handle, why, and what we never touch. We tried to write it in plain language, because privacy policies are worthless when nobody reads them.

1. The short version

  • FameSim simulates a live stream. You are not broadcasting to real people. All viewers, comments, gifts, and leaderboard entries are AI-generated.
  • Your camera and microphone are processed on your device. We never upload raw video or audio.
  • You can choose to sign in with Apple, with email, or as a guest. Whatever you pick is stored in the iOS Keychain — not in plaintext, and never on our servers.
  • To generate AI viewer comments we send a short text transcript of what you say, your chosen mode, and an attested device identifier. We never send video, photos, or personal identifiers.

2. Data stored on your device

  • Account credentials — your Apple Sign-In opaque user ID, email (if you chose email sign-in), or guest UUID. Stored in the iOS Keychain with kSecAttrAccessibleAfterFirstUnlockThisDeviceOnly. Never backed up to iCloud, never synced to other devices.
  • Profile — your handle, display name, avatar photo (if you picked one), presets, fame level, streak, unlocked achievements.
  • Camera feed — processed entirely on your device for the live preview and AR filters. Nothing is uploaded.
  • Microphone audio — converted locally to a short text transcript and an energy level via Apple's on-device speech framework. Raw audio never leaves your device.
  • Recordings — any stream clip you explicitly save is written to your Documents directory or the Photos library. We never see it.
  • Reports & block list — the AI usernames you've blocked and the reports you've filed live only on your device, so your moderation stays with you.

3. Data sent to external services

  • AI comment generation. Each request contains: your chosen username, platform style, personality mode, a ~30-second transcript of recent speech, audio energy level, simulated viewer count, and an attested device identifier issued by Apple's App Attest service. The request is relayed through our Cloudflare Workers proxy to Anthropic (Claude API). The transcript is used only to generate the reply and is not used for advertising or for training any model on our behalf.
  • In-app purchases. Handled entirely by Apple. We receive a signed receipt to verify subscription or coin entitlement. We never see your payment details.
  • Sign in with Apple. Apple shares an opaque user identifier (and optionally your name/email) with us. We never see your real Apple credentials.
  • Crash log. FameSim writes an error log to your device locally. It is not auto-uploaded; you can choose to share it with us when you contact support.

4. What we do NOT collect

  • Your real name, phone number, or location
  • Your contacts, calendar, or health data
  • Your social media accounts, followers, or browsing history
  • Camera video, photos, or raw microphone audio
  • Identifiers for advertising or cross-app tracking (FameSim does not use ATT or IDFA)

5. Trust & safety

  • All comments are produced by AI. We apply prompt-level and output-level safety filters to block harassment, slurs, body-shaming, and sexual content.
  • If FameSim detects language in your speech suggesting self-harm or suicidal ideation, the AI is silenced and an overlay surfaces real-world crisis hotlines (988 in the US, Samaritans in the UK, 182 in Türkiye, findahelpline.com worldwide). We log only that a trigger happened — never the matched text.
  • You can Report any AI comment and Block any AI username from future appearances.

6. Security

  • Account credentials are stored in the iOS Keychain, never in plain UserDefaults, and never copied to iCloud backups.
  • All network traffic uses TLS 1.3.
  • Our LLM proxy requires Apple App Attest on every request, so third parties can't impersonate the app to abuse our AI provider.
  • The LLM proxy holds the third-party API key on our server; the iOS app never sees it.

7. Children

FameSim is rated 12+ on the App Store. We do not knowingly collect personal information from children under 13. If you believe a child has provided information through the app, contact us and we will delete it promptly.

8. Your rights

You can:

  • Export your data — Settings → Privacy & Data → Export my data. Produces a portable JSON file covering your profile, presets, purchases, block list, and recording metadata. You can share it via Mail, Messages, iCloud Drive, or save it locally.
  • Delete your account and all local data — Settings → Delete Account & Data. Immediately wipes your profile, presets, recordings, reports, keychain credentials, and local caches.
  • Manage notifications — iOS Settings → FameSim → Notifications.
  • Manage subscription — iOS Settings → Apple ID → Subscriptions.

FameSim does not track you across apps or websites, so there is no ATT toggle. We honor data-deletion requests immediately and permanently.

9. Data retention

On-device data stays on your device until you delete the app or use the Delete Account flow. AI request payloads are retained briefly by the proxy only for rate-limiting and abuse prevention; they are never associated with your Apple ID, email, or real identity.

10. Third-party services

ServicePurposeData shared
Apple StoreKitIn-app purchasesApple-managed only
Apple Sign-InOptional sign-inApple identifier + (optional) name/email
Apple App AttestAnti-abuse attestationDevice-bound key ID (anonymous)
Anthropic Claude (via our edge proxy)AI fan comment generationLast transcript snippet, mode, platform — anonymous
Cloudflare WorkersEdge proxy, rate limitingAttested device ID + anonymized request body

11. Changes to this policy

If this policy changes materially we will update the "Last updated" date above and, where appropriate, notify you inside the app.

12. Contact

Questions, requests, or complaints: support@appnor.io

We aim to respond within 48 hours.


FameSim is published by Appnor · appnor.io