Voiceprinting in the Age of Artificial Intelligence: From Linguistic Authentication to Linguistic Forgery

Abstract

This study aims to uncover the impact of generative artificial intelligence on the voiceprint in falsifying vocal and linguistic identity, an issue that has become a pressing challenge in the fields of security, linguistic content production, cinematic production, and banking and financial transactions. It does so by reviewing the concept of the voiceprint and its levels, and by focusing on the most significant technical challenges posed by generative artificial intelligence. The study also addresses the principal methods used in detecting voice forgery and cloning. Methodologically, the study relies on the descriptive approach in collecting and analyzing material related to generative artificial intelligence in the domain of voiceprint cloning. The findings show that the application of artificial intelligence constitutes a technical challenge on its darker side, and confirm the importance of addressing these technical challenges through urgent legislation to control this technological revolution, and through the establishment of regulatory frameworks and methods that protect individuals from digital abuses — particularly those related to the hacking of vocal identity, linguistic forgery, and the production and cloning of voices identical to the intended human voice — while also making use of artificial intelligence technologies and applications, with an emphasis on their positive aspects and the avoidance of their negative effects.