The telephon system is pretty much totally broken.
Can you trust incoming numbers? No they can be spoofed. Can you trust the sound of voice? Nope, AI. Can you trust companies to enforce anti robocall and other privacy legislation? Hell no.
US mail isn't much better. The fact that seniors have been scammed by us mail for five decades or more with no improvement to regulation or protection is an embarrassment.
Email of course is a lost cause.
One of the reasons xenophobes are voted into power is the normalisation of attacks, scams and hustles upon your fellow man. If you propell tech forward that bebefits scammers, you are basically distributing leafletts for regressors.
amongst the reasons that the US government spent a bajillion dollars on the world's most secure communications system is to prevent exactly this sort of thing.
the fact that the entire federal government is run by deeply fucking stupid people who are not at all interested in the security of the state is the problem here, for once the AI vendors are not the problem.
the amount of high level government communications being done over public apps and social networks is alarming
even more when you consider highly classified information is sharing the phone with many of those third party apps
I am more concerned with why this just refuses to be funny, is it that humor would take more involvment and knowledge of the details?, is it serious or a fanatical prank?, both?
I hope someone does a late night skit on this and saves me the spit required to figure it out.
Does anyone happen to know this regime's codeword for "the jig is up, time to bug out to China" ? Asking for 350 million friends.
I'm guessing this will lead to some sort of global identification system, so we know for sure someone isn't AI.
When I was in elementary school the police came and fingerprinted us all... you know, for education about the police system or some crap like that. I imagine in the future they'll come and take moulds of your kids vocal chords.
or nothing changes, like with Social Security numbers being inherently insecure and used for reasons outside of their use case, and we just live with that
that already exists, it's called "don't have important conversations on your personal mobile phone, the US government spends tens of billions of years on communications security for this exact reason."
One can dream. Photoshop didn’t inspire a global photo authenticity verification system.
Imagine the decrease in social media volume if users could filter by "real human only, USA."
This would single-handedly solve unemployment!
If this attack worked, voice impressionists would have been used already. I doubt this will be even a little effective except to drum up anti-AI panic, which it will probably do well.
This attack works all the time, look at the "grandma I need $1k to be bailed out of jail" scam.
…without the need for AI, proving my point.
My point is the attack in general works, whether you add AI voice cloning or not as a layer of indirection from the truth.
It doesn't need AI, but it can and does work with AI.
The people in government used to mostly be serious people that took security seriously. It's possible and will likely work because officials are using apps line signal to try and hide the illegal things they are doing.
There overlap between people who sound like Marco Rubio and want to commit espionage is probably small. That is no longer a barrier.