“Any voice they’re speaking is already captured.”
Not that anyone knows.
Shoot, even end-user software is doing that. Certain ones, that is.