US mother gets call from ‘kidnapped daughter’ – but it’s an AI scam | Hacker News Reader

US mother gets call from ‘kidnapped daughter’ – but it’s an AI scam

theguardian.com

50 pointshackernj13 comments

This used to be a common scam in certain developing countries in the 90s and early 2000s.

Scammers would call random numbers until a woman answered and would start by saying "Mom?" and if she said something like "Michael?", they would go "Yes, mom, it's me, Michael, sorry the connection is really bad and something is up with my phone..", because of course their voice would sound different. They would claim to have gotten into a car accident, and that they would be in serious trouble (low rule of law in those countries, problems resolved directly between people) unless they sent those people money right away.

I forget how exactly they expected the money to be sent, maybe it was mobile phone credits or something along those lines.

IIRC the scam was often done by convicts in prison who would get ahold of cell phones somehow and had all the time in the world to try to run that scam.

This exact scam was also performed in developed countries.

The money (cash only for.. reasons..) was collected by a "friend" who would happen to be just right there in the neighborhood.

Very often a car broke down and being stranded far away, no way to pay repairs, credit cards declined for "reasons"..

We called it the "grandmother scam". It does not work with mothers because they tend to recognize their son/daughter voice, even over "bad" connection. But grandma's, aunts and more distant relatives would fall for it.

Unfortunately, if you willingly give cash money to a stranger knocking on your door, there is very little protection/recourse..

This happens often where I live. They call me to tell me my son has been kidnapped.

I just tell them to kill him and stop bothering me… I don’t have a son.

A version of this scam involved SIM-jacking so you'd get a call from the son's/daughter's phone number, and if you hung up and tried to call them you'd either not be able to get through or reach the scammers. I guess with WiFi messaging that's less of a threat nowadays (Hmm, wait, how easy is it to hijack a WhatsApp if you've stolen someone's phone number? Oh, you just login and they'll send you an SMS to authorize? Ignore the "less of a threat" idea then.)

Wild. My neighbor just told me this happened to his daughters a couple of weeks ago too. We're in Arizona as well. One daughter got a call from someone claiming they had her sister and demanded a ransom. They used AI to fake the sister talking. Police were called and they went to the other daughter's home and luckily found everything was fine. The police told them this was a small incident compared to what they had dealt with lately. Sounds like this is starting to happen all over.

At a certain stage the entire process can be automated -- like mass spam emails.

Relevant XKCD: https://xkcd.com/416.

I'm having trouble finding how to correlate XKCD numbers with dates, but I think this might be over a decade old.

ETA: In fact I'm sure it is, because Time (1190) is a decade old this year. Relevant XKCDs: https://xkcd.com/1624/, https://xkcd.com/891/, https://xkcd.com/1686/, etc.

Click the archive link at the top left of the page, then find the comic, mouseover the link to it, and the alt text will be the date it was published.

Even better is just looking at the HTML and searching for the name or the number of the comic instead of playing games with mouseovers and popup text.

XKCD has released thrice a week (monday, wednesday and friday) every week for as long as I can remember. Some back-of-the-napkin math would put #416 in early 2008.

We should start teaching duress passwords at school: https://www.youtube.com/watch?t=54&v=fx77j1vl4d8

The duress phrase my mom and I had was "I love you"

This has been happening a couple of times a month to my elderly father for years, minus the AI. My father has always insisted that "it sounds just like him."

The AI seems like a pointless novelty in this use case, like it was suggested by some middle manager saying "let's see what we can do with AI here. We don't want to get left behind!"

A few years ago, I had a series of spam robocalls. I believe that they were received on my landline which lacked caller ID. So I would pick up and say "Hello", and an elderly woman's voice would say my name, with a question mark. And at first I'd go "Yes?" or "Speaking." and then nothing would happen. I assume that they were fishing for a live human response.

However, it was chilling, because the voice that was selected was very reminiscent of my late grandmother who'd passed away about 20 years ago. And I'd hang up very rattled that some spammer/scammer knew my given name and exactly the sort of voice to get an emotional response out of me.

Reply on news.ycombinator.com