1–6 of 6 posts
Unfortunately, major US financial institutions seem to be ramping up adoption of this technology[2].
Am I missing something?
[1] https://github.com/myshell-ai/OpenVoice
[2] https://www.schwab.com/voice-id
[3] https://www.fidelity.com/security/fidelity-myvoice/overview
TD bank uses your voice as an authentication test. When calling in, if you pass the voice test you don't need to answer further security questions.
voice ID adoption won't work anyways, its not a secure way. consider a person has throat infection, his voice will obviously be deep due to the infection. if we spare that as well, openAI has launched a voice clone AI. Those who are using it now will eventually change it.
Sadly providers' narrow view of "security" fails to recognise that as a lack. It considers it merely an access interruption, workaroundable by fallback to backup methods of authentication, even if that's a costly in-person visit to the bank.
The bigger problem though is the whole idea that something so unsecretable and reproducible as voice should ever be considered for ID.
The sooner voice ID dies, the better, IMHO.