Launch document

Privacy summary

Draft for qualified UK legal/privacy review before broad public launch. This draft is not legal advice or proof of regulatory compliance.

What belongs to you

You keep the rights you already hold in recordings, transcripts, corrections, drafts and exports. Voicesake receives only the limited licence needed to process, store and deliver the service.

What we collect

Account details, project setup, participant consent receipts, recordings you choose to make or upload, transcripts, corrections, drafts, exports, entitlement/payment status and privacy-safe service metrics.

Recording consent

The moderator must give the notice to everyone present and resolve each participant before substantive recording. The organiser remains responsible for applicable recording law; collecting consent in the product does not guarantee universal legal compliance.

Guest contribution links

An invited contributor may continue without creating an account or providing an email address. A guest invitation is limited to one contributor and one project contribution. The browser keeps a private device token so the guest can resume; the service stores only its one-way hash. Anyone who receives an unclaimed invitation link may use it, so organisers should share links carefully.

Retention

Raw meeting audio is scheduled for deletion from active storage 30 days after session finalisation unless the project owner deletes it earlier. Project text, transcripts, approved drafts and existing exports remain available read-only after the project pass expires until the owner deletes the project or account. Provider backup deletion windows must be documented before launch.

AI providers and training

Approved providers may process content to transcribe, organise, translate or draft. Private content is not used for model improvement without a separate, explicit opt-in. Provider retention, regional processing, subprocessors and international transfer terms must be verified before production processing.

During the public beta, automatic live transcription uses the speech recognition service built into a supported browser. The browser vendor may process live speech to return recognised words. Use non-sensitive test speech until that vendor’s privacy, retention and regional processing terms have been approved for production use.

Your controls

Listen during retention, export your content, correct it, delete raw audio early, delete a project and request account deletion. Downloaded copies remain outside the platform’s deletion control.

Support contact is centrally configurable before launch. Product operator: MY BUSINESS OS LTD.