Skip to main content

Privacy Policy

Last updated: December 2025

Zero Data Retention

Utter is designed for zero retention of voice and transcript content. We do not store audio recordings, transcript text, or spoken content after processing is complete.

Our Commitment

Privacy by Design

Only During Dictation

We process your audio only while you are dictating, then delete it when transcription is done.

No Model Training on User Content

Voice and transcript content are not used to train our models or create behavioral profiles.

Limited Service Diagnostics

We collect limited technical diagnostics (for example, app version and crash reports) to maintain reliability.

How Utter Works

When you initiate dictation, audio is transmitted over encrypted channels to the selected transcription provider, converted to text, and returned to your device. Processing is session-scoped and does not create a retained archive of your audio or transcript content on Utter servers.

Information We May Collect

To operate and maintain the service, we may collect limited technical diagnostics, including:

  • Device type and operating system version (for compatibility)
  • App version and crash reports (for debugging and improvements)

These diagnostics are used for compatibility, reliability, and troubleshooting, and do not include dictated content.

Security

We apply administrative and technical safeguards appropriate to the data we process, including encryption in transit and access controls for supporting systems.

Service Providers

Third-Party Services

Utter relies on third-party processors for transcription and optional AI features. Data is shared with these processors only as needed to fulfill the request you initiate.

Transcription Services

We use OpenAI by default, with an option to select Deepgram, for speech-to-text processing.

AI Processing Services

If you enable optional AI features (such as formatting, rewriting, or AI modes), transcript text may be sent to the provider selected in your settings:

  • Anthropic (Claude)
  • OpenAI (GPT models)
  • Google (Gemini)

Important Notice

Each service provider maintains its own privacy policy and data-handling terms. You should review those policies to understand provider-specific practices.

Where supported, we configure integrations to opt out of data retention and model training on user content. Provider policies may change over time, so you should periodically verify current terms.

Changes to This Privacy Policy

We may update this Privacy Policy from time to time. Any update will be reflected by revising the "Last updated" date at the top of this page. Your continued use of Utter after an update means you accept the revised policy.

Contact Us

If you have questions about this Privacy Policy or our data practices, contact us through the Help Center or official support channels.