How to stop Abridge training AI on your data
What the policy itself says, plan by plan. Where Abridge publishes the steps, they're here in its own words, read from an archived copy of the page that gives them. Where it doesn't, this page says so instead of inventing a menu path.
All accounts
Abridge's privacy policy contains no form of "train" and no machine-learning or artificial-intelligence language, on 14 Aug and again in the revision dated 27 Aug 2026. It transcribes and summarises medical consultations, which makes the input clinical conversations between doctors and patients. The August revision adds a biometric section that comes closest to the question: voiceprints, recorded voluntarily, are used "solely to improve the accuracy of dictation, speech recognition, speaker identification, and transcription" in your own sessions, with consent, a one-year retention limit and destruction on request. That governs the enrolment recording, not the consultations themselves — on whether those train a model, the document is still silent. There is a structural reason for that silence, and a careful reader meets it on the first screen. The policy's applicability section says it "does not apply to the content that our customers (e.g., hospitals, healthcare systems) and their authorized users upload and store in our services", that Abridge processes that Customer Data "solely on behalf of our customers", and that a patient should "refer to your provider's Notice of Privacy Practices". The consultations are governed by contracts with hospitals and by each provider's own notice — documents this site cannot read. The absence is by design rather than by accident, which is a different finding from silence, and the entry now says which one it is.
Did a step here not work, or does Abridge offer a control this page doesn't mention? Tell us and we'll correct it.