1. Summary
Bean is an iOS app that lets you talk to an AI assistant that can see your screen. It is designed so that we collect nothing about you and save nothing about your sessions.
- No account. No login. No email collected.
- No analytics tied to you. No advertising IDs. No third-party trackers.
- Your conversations are not stored on our servers or on your device.
- Audio and screen frames are streamed to our AI provider (xAI) for the live answer, then immediately discarded.
If you stop using Bean, there is nothing to delete because there is nothing about you to delete.
2. What data we handle
Bean handles the following categories of data, all of them in transit only:
Audio
When a session is active, your microphone audio is streamed to xAI so the assistant can hear what you are saying. Audio is not recorded to disk by Bean, is not retained by xAI under their commercial API terms, and is dropped as soon as the response is generated.
Screen frames
When you start a screen broadcast, iOS sends frames of your screen to Bean's broadcast extension, which forwards them to xAI in real time. Frames are processed by the model and then discarded. They are not saved as files, screenshots, or video.
Model responses
Text and synthesized speech generated by the model are returned to your device and played back live. We do not log responses on our servers.
No personal identifiers
Bean does not ask for, collect, or transmit your name, email, phone number, address, Apple ID, device ID, IDFA, IDFV, or any other identifier that would tie a session to you specifically.
3. What happens in a session
The lifecycle of a Bean session is:
- You launch Bean (via the Action Button, Back Tap, or the app icon).
- iOS presents a system Start Broadcast sheet. You tap Start.
- Your microphone audio and screen frames begin streaming to xAI.
- The model generates a spoken response and plays it back.
- You end the session by tapping End or by stopping the broadcast from iOS.
- The stream closes. Nothing about the session is written to disk or to our servers.
Bean does not keep a transcript of your session, either on your device or on our servers. If you want a record of an answer, copy it yourself before ending the session.
4. xAI as a subprocessor
Bean is a thin client. The actual model that understands your screen and generates the response is operated by xAI Corp., accessed through their commercial API.
Under our agreement with xAI:
- Inputs you send (audio, screen frames, text) are not used to train models.
- Inputs are processed for the duration of the request and are not retained for analytics, profiling, or any other secondary purpose.
- xAI is SOC 2 Type 2 audited.
xAI's own privacy practices for its API are published at x.ai/legal/privacy-policy.
We don't have a separate "training data" or "improvement" toggle in Bean because we never opt you in. Your inputs are never used to improve the model.
5. Data on your device
Bean stores a small amount of local-only data on your iPhone using standard iOS preferences storage. This stays on your device and is never transmitted to us:
- Your selected voice (which TTS voice the assistant uses).
- Whether you have completed onboarding.
- Permission status flags (whether you granted microphone and screen recording).
Deleting Bean from your iPhone removes all local data.
6. No tracking or advertising
Bean does not include any third-party analytics, advertising, or attribution SDKs. We do not embed Google Analytics, Firebase, Amplitude, Mixpanel, Meta, AppsFlyer, Adjust, or any equivalent.
We do not buy or sell advertising and we do not participate in any cross-app or cross-site tracking.
For the App Privacy disclosure in App Store Connect, audio and screen content sent to xAI during a session are declared under App Functionality. They are not linked to your identity and are not used to track you. No other data categories are collected.
Crash reports, if any, are limited to Apple's built-in opt-in diagnostic system. Apple controls that flow; we receive only aggregated, de-identified crash logs through App Store Connect when users have opted in via iOS Settings.
7. Children
Bean is not directed to children under 13 (or the equivalent minimum age in your jurisdiction). We do not knowingly collect any data from children. Because Bean does not collect personal data from anyone, this is enforced by design rather than by policy alone.
8. Your rights
Privacy laws like the GDPR (EU/UK) and the CCPA/CPRA (California) give you rights to access, correct, delete, and port personal data that a company holds about you. Because Bean does not hold any personal data about you:
- There is nothing for us to provide in response to an access request.
- There is nothing for us to correct.
- There is nothing for us to delete.
- There is nothing for us to port.
If you believe we have somehow handled data about you and want to confirm, you can write to us at the address below and we will respond within 30 days.
9. Changes to this policy
We may update this policy from time to time, for example if we change a subprocessor or add a new feature. When we do, we will update the "Last updated" date at the top of this page. Material changes will also be announced inside the app on next launch.
Because we do not have your email, we cannot notify you individually.
10. Contact
Questions about this policy or about how Bean handles data: