UNESCO Room 9
7, Place Fontenoy
75007 – Paris
11:30 a.m. – 12:30 p.m.
Artificial intelligence and voice identity theft
As we enter a society of verisimilitude, synthetic voices and neural models make the false almost indistinguishable from the true. Driven by new digital uses, mobility, the explosion of podcasts and the rise of text-to-speech, audio is experiencing unprecedented growth and becoming a global information vector, sometimes essential in countries with poor access to the written word.
But this revolution is accompanied by major risks of usurpation, whether personal, economic or political. The ability to reproduce a voice to perfection opens the way to manipulation, fraud and new forms of psychological influence, as specialists in the human, cognitive and digital sciences have pointed out. Similarly, artificial intelligence is having an impact on a number of business sectors where the human voice seems threatened by a gradual trend towards substitution, notably in the cultural and creative industries ecosystem.
Faced with these crucial challenges, it is becoming essential to establish an ethical and regulatory framework. Several international institutions have already begun this process, but collective vigilance remains essential to ensure that technological progress remains at the service of human beings.
The resolution “Artificial intelligence and voice identity theft”, supported by UNESCO’s Sound Week with several international delegations, is part of this process, establishing an inventory of the current situation, a technical and legal analysis, followed by concrete proposals based on expert debates.
With this in mind, three debates will be organized in 2026: the first on the general state of play, the second on legal solutions, and the third on technological solutions.
First debate with:
– Nicolas Curien, member of the Comité consultatif national d’éthique du numérique
– Laurence Devillers, AI professor at Sorbonne University, CNRS researcher at the Laboratoire interdisciplinaire des sciences du numérique (LISN) on human-machine affective interactions.
– Erik Orsenna, writer and academician
– Jérôme Doncieux, founder and CEO of ETX Majelan
– AI expert, United Arab Emirates
Moderator: Christophe Rioux, academic, journalist, writer and administrator of UNESCO Sound Week
