This paper explores the feasibility of identifying authentic source speakers from voice conversion-generated utterances, which pose potential risks for deception and privacy violations.
