OpenAI Cautiously Releases Powerful Voice Cloning Tool
-
OpenAI built a voice cloning tool called Voice Engine that can create a synthetic copy of someone's voice from a 15-second sample. However, there is no public release date yet.
-
Voice Engine is built on the same model that powers the voices in ChatGPT and OpenAI's existing text-to-speech API. It's priced at $15 per 1 million characters.
-
OpenAI is being cautious about the release to prevent misuse, only giving access to around 10 developers working on "low risk" and "socially beneficial" use cases.
-
Safeguards include watermarking clones, providing access to red team security experts, and testing identity verification measures.
-
The timeline for a wider release depends on what OpenAI learns from the initial pilot about safety issues and preventing confusion between real and synthetic voices.