No, it’s not a thing.
First of all, no contract that I’ve ever heard of is bound by an audio file of someone’s voice.
Second, I’m not an expert, but I’m willing to be any audio system can tell the difference between a live (analogue) voice and a digital recording, even if your ears can’t.
Third, even if a voice print was used as a binding contract, how would anyone know it’s YOURS? There’s no Registry of Voices.
My point was - if a company claimed you had agreed to a purchase of their magic elixer and billed you or something or somehow got your credit card, by the time the consumer bureau got 100 complaints about the same fraud from the same company, I assume they would be in a bit of legal trouble. The credit card company would also have something to say to them. The “switch long distance providers” I assume worked because you pay your phone company, and they forward the costs for the long distance charges to the company you are allegedly signed up for. So unless you read your bill carefully or noticed an uncalled for increase in long distance charges, nobody notices… Unless the long distance company investigates why so many customers are ditching them. I assume there are similar scams over subscriptions, but the key is you have to not notice your payments have changed to someone else.
I tend to either wait, on unrecognised numbers, to listen to what they are selling, if I even bother answering.
The word “What” is more used than “Yes”; the words “drug addict” and “bankruptcy” often feature if I am in a good mood.
I try not to troll, but… I cannot help myself on occasion with these assholes.
It helps that my country specifically has made cold calls illegal, so after the unfortunate early stages I can quote the specific law that must make them stop, so the phrase “section 45(b) of the South African Electronic Communications and Transactions Act” is way more likely to come out my mouth than the word “yes”.
I so very rarely get similar calls that I think I am on an international (or maybe just Nigerian) blacklist for being a terrible, terrible “customer” experience.
I’m not sure how to place this statement. Are you talking about something like the 1970s? That would have been the beginning of digital recording technology. Likely even your ears would have been able to tell the difference between live and recorded, as well as between digital and analogue recording.
The supposed fraud back then would certainly have used analogue recording only. It would have used an analogue recording of you saying yes “live”. However, because the entire scam would have relied on a recording, there was nothing to be distinguished by any supposed technology.
In our modern era everything, of course, is a digital recording, including the voice of the person you are speaking to “live” on the phone.
This is one of those mild fears that nags at me. There are worse fears keeping me up at night, but this is still an unpleasant one.
You see, I record a ton of YouTube videos, and I have never made a serious attempt at anonymity–for example videos of my wife and I playing as a piano/bass combo always have both of our names.
So in my photography videos there are hours and hours of me saying so much stuff that AI would be able to make a perfect me.
On a positive note, I run my mic through a channel strip to do standard voice processing (compression + desser + expander + parametric EQ) so the nice rich radio voice that AI will generate will probably not match my natural voice.
Its how the trick old people into sending money overseas to a young relative in trouble. Pretty sure this is how they scammed my elderly aunt out of $5K by someone sounding just like my brother claiming to be in the hospital in Barcelona.
Yep. And that’s why I tell everyone I can about “proof of life” questions. That’s those kinds of questions that Vietnam-era pilots had to provide on a file card before going into combat, things like “What car did you work on with Kyle last summer?” that nobody would ever figure out.
I have already told my dad that if he ever hears me with a tale of woe on the phone, he ought to be asking questions about what we had for lunch last time we met and similar things.
ETA: Forgot to mention, the purpose of the note cards for the pilots was so that if they were ever shot down and calling for help on the radio the rescuers could prove they were who they said they were and not the enemy impersonating them.
So, two factor authentication, invented before the internet.
we also have a cold-call law (must be coming from a 809 prefix - which are easily suppressed and dont even ring on my phone anymore) …
as we say here: ¡Santo remedio! … (holy remedy - “does the trick”)
We have a similar system, but with set of non-sequiter responses that signal “all is well” or “this is really me”. Any normal-sounding response signals a problem.
As far as spam/scumbag calls, we only answer a specific and very small set of phone numbers. Everything else goes to voice mail. I haven’t talked to a spammer in years.
I have just recently received a run of robocalls where the first thing the robot asked was ‘can you hear me OK’ and it made me think… I still don’t believe the myth is true, but…
I find it somewhat plausible that someone on the spectrum between unscrupulous salesperson and scammer (not a marathon distance), rather than actually trying to legally enforce something purporting to be a binding contract, might merely apply pressure by arguing something like “Look you have to pay for our service, because I have a recording of you saying yes. If you don’t, I will have to refer your case to our debt recovery department”.
I think they’re just trying to engage you.
Probably.
If there is to be a modern version of this myth, it needs to be updated to ‘don’t say anything’ - because the technology exists for a bad actor to record a relatively short sample of speech and use that to train a generative algorithm to output realistic synthesised speech that appears to be saying anything they like, including ‘yes’.
Edit: the thing that makes this less likely to actually happen is that we’re being asked to believe that scammers would pursue a false contract all the way to court. That would mean coming out from under their rock, exposing their identity, making themselves known to the law.
Why would scammers from outside the country have to worry about our courts?
Probably not, but the whole notion of their being a legally-binding contract based on a snippet of audio, rests upon the idea that the (dishonest) party who wishes to enforce such a contract, will be prepared to resort to legal means to uphold their dishonesty. It’s fundamentally self-contradictory.
I’m not holding my breath, at least not in the near term, since the scripted greeting on the other end, even if it is AI generated, is just as stilted and laggy as older pre-recorded greetings or the delay between answering the phone and their system connecting to a live scamm…I mean operator. It’s no less obvious with AI than it was in the past, the script is just different.
When previously I’d answer the phone “Hello Company X” and the other end was silence…silence…silence BWOOP “[heavy foreign accent] Hello may I please speak to the person responsible for your office cleaning?” Now I answer the phone with “Hello Company X” and the other end is silence…silence…silence “[obviously AI voice] Are you there?” or “Don’t hang up.” That’s when I hang up.
I wonder if this could be by design, similar to classic Nigerian prince and other scam emails or text messages. They are purposefully suspicious and transparent because any one who would fall for something so obvious is a good mark. If they were more clever in their deception they’d waste time and resources going further down the process of the scam before the mark catches on, often after a bunch of live phone calls and/or email exchanges. Better to weed out the credulous first and not waste time on them.
In about 2021, some wires got crossed at State Farm (my home insurer) and I got briefly mistaken for someone who had a big house or garage loss, fire or tornado or flood from the sounds of it. It got fixed but not before my info got turned over for referrals. I started getting contacted by contractors and lawyers (and Dick and Harry) and I had to quickly get them to go away and stop trying because “You have the wrong person” isn’t enough. I started taking on a third person secretarial calltaker persona if they get pushy.
Caller: Is this Name?
Hero: Who is calling?
C: mumbles something quasiofficial, is this Name?
H: What’s the purpose of your call?
C: It’s an important message that can only be given to Name. This is him, isn’t it?
H: This is Name’s duly appointed representative. Name screens their calls and understandably disregards those without a message.
C: Uh-huh. Well, tell NAME that we’re offering up to 45% off to new, highly-qualified customers …
I’m not talking about them generating realtime AI responses - I mean that the technology already exists to record your voice saying “No, I am not interested please go away and die in a fire”, and from that voice sample, generate a plausible-sounding synthetic recording of you appearing to say “Yes, please sign me up!”. ‘Don’t say yes’ wouldn’t be enough, if this scam was real.
The technology exists to do that now, but the reason this scam still doesn’t seem to actually exist is that legally enforcing a contract means taking it to court and scammers aren’t going to want to do that.
I assume too, they could take an incorrect response as “I am doing this under coercion”.
The problem too is that it exposes a pattern if they end up having to prove something in court, or to a credit card company, too often. To quote Goldfinger, “once is happenstance, twice is coincidence. Three times is enemy action.”