AI Voices Are Now Indistinguishable From Humans, Experts Say
SMRTR summary
A groundbreaking study reveals that AI-generated voices are so sophisticated that people can only distinguish them from real human voices about 60% of the time, effectively passing the audio Turing test. This poses significant risks to financial security, as many banks rely on voice identification for customer verification and transaction authorization. OpenAI's CEO has warned that AI has already defeated most bank voice ID systems, while fraud experts note that scammers are increasingly using voice-cloning technologies for personalized financial attacks.
SMRTR provides this summary for quick context. The original article belongs to PYMNTS.
Read the original article