I assume that if somebody had physically lost their ability to speak, it would now be possible to generate a pretty reasonable synthetic voice. Should we all be

I assume that if somebody had physically lost their ability to speak, it would now be possible to generate a pretty reasonable synthetic voice. Should we all be archiving a high-quality voice sample as insurance? The implications for security are huge. If your friend calls you up for a very quick chat from an unknown number and asks you to remind them of your address, are you going to ask for authentication to prove it's really them and not a convincing synthetic voice?