Clarification on Voice Cloning and Licensing for Chirp3 / TTS Voices

Google Cloud Build with AI AI APIs
1

Hello Team,

I am currently evaluating Google Cloud Text-to-Speech, specifically the Chirp3 and other prebuilt voices. I would like to clarify the licensing and usage terms:

  1. Am I allowed to use TTS-generated audio (e.g., Chirp3 voices) as training data to create a cloned/custom voice in my own model or another provider’s system?

  2. Could you confirm that the license for prebuilt voices prohibits extracting or replicating them outside of Google Cloud?

  3. Does it make sense, from a technical or compliance perspective, to use Google TTS-generated audio to create a custom voice within Google cloud?

Thank you for clarifying — having this confirmed will help us ensure compliance with Google’s licensing.

Best Regards