Voice cloning learns a target voice from recordings and then synthesises new speech in it. It enables personalised dubbing and accessibility, but also impersonation, so consent and disclosure rules — including the NO FAKES Act — govern responsible use.