Back to home

Privacy

Last updated: September 27, 2026

Your records and creations are private by default. Here’s how we use your data, and how you can manage it.

Manage your data

Manage your choices

Manage consent for responses & organization, export data or delete your account in Settings → Data & privacy. Manage personal voice in Settings → Voice → My voice.

Turning off responses & organization does not turn off every AI feature. Speech transcription, photo responses, creation and personal voice follow their separate permissions.

What We Process

Content you provide: text, recordings, photos, and additions or feedback related to your records.

Content made for you: responses, conversations, organized records, and creations such as images, music and lyrics.

Usage information: actions such as saving and sharing, sign-in method, account ID, email, entitlements, device and app version, and necessary security records.

Service regions and data storage

Chinese mainland service

When you choose the Chinese mainland service, your account, records, creations and related usage data are stored and processed on servers in the Chinese mainland. Features such as transcription, AI responses, creation and personal voice use third-party models within the Chinese mainland, without forwarding processing to the international service.

International service

When you choose the international service, your account and content are hosted on servers in Singapore. Some AI features use third-party services in other regions, as described below.

Separate accounts and content

Accounts and content are managed separately in each region and do not sync automatically. Using a Chinese or English interface does not change your service region or move existing content.

Apple services such as installation and sign-in, and external links you choose to open, follow the respective providers’ rules.

Records and AI features

Voice records and original audio

When you record and agree to cloud transcription, the recording is uploaded to a third-party speech-recognition service and converted into text you can check and edit. Transcription and keeping original audio are separate: original audio is saved privately with a record only after you explicitly agree the first time. You can delete individual audio segments in the record details while keeping its text and response.

Original audio kept as a record attachment is not used by default for responses, long-term organization, model training, voice cloning, or sharing. Responses and organization process the record text and context needed for those features.

Responses & organization (Optional)

After you explicitly agree to “Responses & organization” in the app, record text needed for responses and organization is sent to third-party AI services. When automatic organization is available, BYOME processes eligible new records saved after you agree in the background, without asking again for each record. Inputs and generated outputs may be used to improve model services under the provider's applicable terms.

You can save records without agreeing. Automatic organization requires your explicit consent and does not automatically include records saved before you agreed. This permission does not replace separate choices for photo responses, creation or personal voice.

You can withdraw this permission in Settings → Data & Privacy. After withdrawal succeeds, no new response or automatic-organization processing is started and newly returned results are not saved; existing records and organization results remain available. Requests already sent to a third party cannot be recalled. Agreeing again allows only subsequently saved records into automatic organization.

Photo responses (optional)

When photo responses are available, we send your photo and added text to a third-party AI service in China only when you choose Get a Response. Information attached to the photo is removed before sending. Separately saved location data is not sent, and other records are not read.

Under that service's current public terms, submitted inputs and generated outputs may be used to improve model services. Its applicable rules govern retention, and BYOME cannot currently promise immediate third-party deletion or a fixed third-party retention period.

Deleting the photo from BYOME removes the files and linked responses managed by BYOME, but cannot undo processing already completed by the third-party service. Not tapping “Get a Response” does not make this third-party photo-understanding request. Photo understanding is not used for face identification or sensitive-attribute inference, and generated content does not automatically become a fact about you.

World Topic Dialogue

When you start a topic conversation, its question, your messages and the necessary context are sent to third-party AI services to generate and review a response. Other records are not read, and AI responses are not treated as facts about you. With personal voice enabled, response text is used to generate audio only when you tap play.

Recent Record Revisits

To help you revisit recent experiences, we may process a small set of recent record text and related responses in the background. The resulting reflections are private to you. You can view their sources, delete them individually, or export or delete them with your account data.

Creations and personal voice

Creation (optional)

When creation is available to your account and you select records and confirm creation, BYOME sends the necessary text from those records to third-party AI services to assess, plan, and review an image or music work. The music-generation service receives the music prompt and lyrics needed to generate audio, not your complete source records. Lyrics and images may still express private experiences. A music work can include a cover image. Saving a record or browsing existing works does not start this creation request.

Third-party data rules may differ. Text-planning inputs and outputs may be used to improve model services under the applicable terms. Creation in the Chinese mainland service uses models within the Chinese mainland. For the international service, image generation is accessed through Singapore, where data at rest is stored; inference may run elsewhere outside the Chinese mainland. Music generation may process data in the United States or other regions. BYOME cannot promise a shared retention period or immediate deletion by every provider. Deleting content from BYOME cannot undo completed third-party processing.

Creations are private by default and can be deleted individually. Images and lyrics may reflect private experiences. Deleting a source record does not delete or rewrite a completed work; delete the work too if you want to remove it. Creations are generated expressions, not facts, psychological assessments or predictions.

Personal voice (Optional)

When the feature is available to your account, you can create a personal voice in Settings → Voice → My voice. Only after your separate agreement does BYOME record and upload a dedicated reading to create a synthetic voice for your account; it does not read other recordings on your device.

Recordings, voice identity data and text for speech are processed by the voice provider for your service region: in the Chinese mainland for the mainland service, or in Singapore for the international service, for private playback. The dedicated recording is not made public or used to train general-purpose models; we delete the temporary recording after creation and verification. Generated audio is kept for at most 1 day before entering deletion. You can disable or delete your personal voice at any time. Deletion covers related voice data and generated audio held by us and the third-party service, and is complete only after cleanup is verified. De-identified deletion evidence is retained in the online system for at most 30 days.

A synthetic response does not mean you said it. Deleting personal voice does not delete ordinary records, their text or everyday original audio.

Sharing and widgets

Music Share Links (Optional)

A public listening link is created only when you choose Share full audio. Anyone with the link can listen and read the lyrics, but cannot see your source records. The page is not indexed by search engines, but the link can still be forwarded. You can stop sharing at any time; deleting the work or account also disables the link.

Home Screen widgets (Optional)

After you add a widget, content you allow or pin appears on your device’s Home Screen and may be seen by others nearby. Automatic display and pinned content are managed separately; turning off automatic display does not remove pinned content. Widgets do not publish content online, generate new works or play music automatically.

Storage, Export and Deletion

Export and deletion

You can export data or delete your account in Settings → Data & privacy. The JSON export includes records, responses and information about works. It does not include image or audio files and cannot be used to restore your account.

Deleting a record also removes its attachments and responses. Completed works remain; delete them separately if you want to remove them. Deleting a work keeps its source records and stops public sharing of that work.

Once account deletion is accepted, you are signed out and can check progress in the app. Deletion applies to your current service region and cannot be undone once complete. Copies already downloaded or shared with others cannot be recalled.

Backups and recovery

Backups are only for service recovery, not personal data restoration. Deleting an account does not remove all backups immediately. We apply existing deletion requests before restoring service.

International-service backups follow a 14-day retention window; cleanup of some older copies may take longer. Automatic expiry cleanup is not yet enabled for mainland-service backups, so those copies remain.

Usage measurement and website

Product usage measurement

We collect feature-use events linked to your account, such as opening the app, viewing a response or interacting with a work, to understand use and improve the experience. Data includes action times, app version, installation channel, session and content identifiers, but not record text, photos, audio, lyrics or private links. It is not used for cross-app advertising tracking. Data is processed by servers in your current service region.

Pending events stay on your device for at most 7 days and are cleared on sign-out or account deletion. Raw server events are kept for at most 90 days and can be exported or deleted with your account. These limits do not apply to records, support correspondence or security diagnostics.

Aggregate results without account, session or content identifiers or individual activity details are not removed when an individual account is deleted.

Visiting the website

Launch-notification registration is closed. The website no longer accepts new signups. Previously registered email addresses, registration times, sources and browser information do not create an app account and are not used for product usage measurement. Contact support@byome.online to cancel a registration or delete this information.

The website remembers your light or dark theme choice in local storage in the current browser for future visits. You can clear this preference through your browser.

Contact us

This notice covers the BYOME app and website provided by Shanghai Guangyin Zhinian Technology Co., Ltd. Contact support@byome.online about data processing, consent, export or deletion.

Back to top