Fact · Dictation software
01 = yes, 0 = no
Diction's privacy policy says its audio and transcriptions are never used to train, fine-tune or improve any models.
Quote found word for word on the live source, last checked October 20, 2026
In the source's own words
“Your audio and transcriptions are never used to train, fine-tune, or improve any models.”
Diction privacy policy, Diction Labs, 2026, Section 'Diction One (Cloud)', 'No model training.'.
Open the original- What it measures
- The policy states this for the hosted Diction One service. Its on-device section says audio never leaves the phone.
- What it does not show
- The statement is in the Diction One section. It does not cover self-hosted servers, which the user runs.
In the data
A row of Dictation App Features Compared: Offline, Privacy, AI, as Diction: training on audio: 0 1 = yes, 0 = no.
- App
- Diction
- Feature
- Recordings can train the maker's models
- Detail
- Diction One cloud section
- As of
- 2026-10
- Note
- 0 means the maker's page says the app does not do this.