The percentage (77% of victims lose money) is totally unfounded.

The original report by McAfee states that a market research company has asked people in a survey whether they would give money if a loved one asked for it. Even if you value this kind of research, it does not at all imply that people are falling for current AI voice scams.

> The global average showed 25% of people surveyed either experienced an AI scam or someone they know did.

And the article lost all credibility.

I received a voicemail customized to me, referring to my company name and state, that sounded like a friendly, informal older gentlemen with a bit of a southern accent. It seemed to probably be selling something so I didn't take it too seriously. But only when he said the callback number - "Plus one, two oh two, ..." did I realize it was a computer-generated message. Not a scam, just marketing, but it was scary how real the voice sounded.

Questionable statistics aside, it's unbelievable to me that the US hasn't been able to greatly reduce phone and SMS scams. In my opinion, platforms like Twilio should have some liability for scams made using their platform, and we should put some serious economic pressure on countries that aren't cracking down on scam organizations within their borders.

Doesn't this require a huge amount of voice data before it would even sound remotely believable to maybe an elderly person with hearing/processing issues? I thought this was only possible for YouTubers,news presenters, etc as they have many hours of voice data to use.

Voice has to be the low hanging fruit. Even if the article is not credible, I am sure we will ramp up towards this sooner than we expect