Account and settings

How Talk & Comment processes transcripts and AI features

Learn how Talk & Comment processes transcripts and AI features, which account or school settings apply, and where to find privacy details.

Last updated October 16, 2026

Use this guide to understand where enabled AI features send audio and text, and which account and school controls apply.

Where AI processing happens

AI processing uses Amazon Web Services, Inc. Audio and text are sent to AWS to produce transcripts and, where enabled, summaries, suggested titles, and translations you request.

For a school or district with a signed agreement that sets a later start date, the change takes effect on the date in that agreement. After the applicable cutover, Talk & Comment-hosted processing is no longer available, even as a fallback or a saved preference. If cloud processing is unavailable, the request does not switch to Talk & Comment-hosted processing.

The Privacy Policy and Sub-Processor List describe AWS processing in the United States, input retention, and the commitment not to use AI inputs to train or improve models.

Which features use AI

Feature What it does
Transcription Creates text from a recording's audio.
Summaries and suggested titles Generate text from a transcript when enabled and permitted.
Translation Translates transcript text into a requested language. The original audio stays unchanged.

Read along displays an existing visible transcript while the audio plays. It is not a separate AI-generation request.

Transcript generation, summaries, suggested titles, and translation require TAC Plus or higher entitlement, including eligible School & District seats. A paid plan does not override school AI restrictions. Listeners do not need a paid plan to read a visible shared transcript or use translation that the author allows.

Account settings and queued requests

In Settings, Processing identifies the provider. It does not offer a switch to Talk & Comment-hosted processing after cutover. An older saved processing choice does not change the current route.

Auto-transcribe new voice notes controls whether new uploads queue transcription automatically. Turning it off stops automatic requests for future uploads; it does not recall a request already sent to the provider.

Automatically use suggested titles and Let viewers translate my notes are separate settings. Viewer translation starts off and applies only to transcripts you have made visible.

Talk & Comment checks current entitlements and school AI permissions again when queued transcription work begins. Work may be skipped if the account is no longer eligible.

School and student controls

A school's AI transcription policy controls whether its members can use transcription and related AI features. Recording permission is separate from AI permission.

Student and under-13 accounts require organization authorization for AI. Without it, no AI processing occurs for those accounts. A student's own choice or a guardian's consent record does not independently enable AI.

Student AI settings determine whether transcripts and translation are allowed, and whether summaries and suggested titles are allowed too. To change AI for students, a school or district administrator must contact [email protected] with a written district request. This is not a self-service student-AI switch in the school admin console.

For the available controls, read the school admin console guide. For recording, review, and listener steps, read transcripts and translation.

Privacy questions

Read the Privacy Policy and Sub-Processor List, or email [email protected]. Schools can also review the privacy and DPA guide.

Was this article helpful?

Still stuck? Email support and a real person will get back to you.