Re:Fraudsters Used AI to Mimic CEO’s Voice
by hanniabu on Tuesday September 03, 2019 @07:18AM
(#20865415)
For records I'm sure the infer call approvals are recorded and then after a dispute they went back to listen and realized there that it did sound electronically produced, but those enough to where you don't notice if you're not looking for it.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by Traster on Tuesday September 03, 2019 @05:55AM
(#20865064)
If your level of security is "This guy on the phone sure sounds like someone I should wire several hundred thousand dollars for" you should probably look at improving your security procedures.
Whilst it's funny these guys did this with AI I can't help but feel that they could have done this with someone who is halfway decent at accents and saved themselves a lot of effort.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by dgudkov on Tuesday September 03, 2019 @07:02AM
(#20865354)
It's just a matter of scale. For some companies $250K is pocket lint.
Would you tell your wife your credit card number if she called you and told you her card is blocked for some reason and she needs to refill the gas tank or buy something?
Re:Fraudsters Used AI to Mimic CEO’s Voice
by Tade0 on Tuesday September 03, 2019 @07:39AM
(#20865537)
A while ago a friend who I hadn't seen since 2011 messaged me on Facebook asking if I could take a look at his account on a guitar forum we both used to frequent, because he thought his account was stolen, but couldn't check it for activity without making a new one(and probably also having enough posts to be considered "in").
So I asked who sold him that Music Man Petrucci 7-string in 2009(or so) and after he answered correctly, I followed through with his request.
My point being: it helps to have such a question prepared for everyone - or alternatively - a question concerning some fact about you that only people who know you could answer.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by Traster on Wednesday September 04, 2019 @05:47AM
(#20875521)
Funnily, companies for whom $250k is pocket lint since they're big enough to have processes for payments and it's often incredibly difficult to circumvent those processes.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by phpnode on Tuesday September 03, 2019 @06:23AM
(#20865179)
1. Yes, the ease with which attackers can take over a phone number is a big security issue in general
2. You might question your wife's request but people are far less likely to question their CEO in the same way, it might be seen as insubordination
Re:Fraudsters Used AI to Mimic CEO’s Voice
by johnnycab on Tuesday September 03, 2019 @07:29AM
(#20865486)
>Whilst it's funny these guys did this with AI I can't help but feel that they could have done this with someone who is halfway decent at accents and saved themselves a lot of effort.
On the contrary, it seems like they did save themselves a lot of effort by deploying the de-centralised 'deepfake' AI model. It would be significantly hard for the opportunistic scammers to go out talent-scouting locally, trying to match the exact vocal characteristics of a person. It would also create a vulnerability, notably the person hired to impersonate the CEO.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by ThePadawan on Tuesday September 03, 2019 @06:40AM
(#20865254)
Woah there. Maybe I'm way off base here, but I wouldn't even come in to work overtime on a weekend without written confirmation from my boss, never mind touch anything money-related.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by taneq on Tuesday September 03, 2019 @07:14AM
(#20865403)
Imagine getting an email from your boss asking you to order in some supplies from a new supplier. You think it sounds pfishy (hah) and so you call them to confirm. They confirm.
Do you really think that most employees would need more than that?
Re:Fraudsters Used AI to Mimic CEO’s Voice
by ThePadawan on Tuesday September 03, 2019 @08:17AM
(#20865777)
I might be biased since I live in a country famous for its amount of bureaucracy (Switzerland).
But yes, at my current employer, I believe no one would do that just based on an email and a phone call. But then I dare say that faking Swiss German is an order of magnitude harder than faking English.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by michaelt on Tuesday September 03, 2019 @07:32AM
(#20865500)
One of the most common types of this fraud is to have hacked the CEO's e-mail account. In which case not only is the request written, it also has the expected from address, footer, DKIM and so on.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by ConfusedDog on Tuesday September 03, 2019 @07:07AM
(#20865374)
Deepfake voice is just making social engineering more available to non-performing-artist type of attackers. Spear phishing is just getting to a new level. This type of things gonna happen more to 'public' figures.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by xiphias2 on Tuesday September 03, 2019 @05:50AM
(#20865040)
Right now I don't know if banks offer the CEO to approve big transactions easily on his mobile phone, but that would be a possible solution (email is not structured enough for this).
Also while Bitcoin community is creating new standards for collaboration, like multisig transactions and partially signed transaction format (which would help here in creating a bank-independent solution), I don't see the same standardization happening in the banking system.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by TheDong on Tuesday September 03, 2019 @06:59AM
(#20865334)
Your definition is a little off.
A computer capable of thinking like a human is "strong ai", or "general AI", or any of a few other poorly defined terms.
The turing test is a poor approximation of that goal. Computers can and will pass the turing test long before they're actually capable of real human thought.
Specifically, the turing test is merely having a human judge evaluate a free-form text-based conversation with a partner, and them being unable to distinguish a computer from a human in that scenario.
The main thing this instance lacks is the free-form nature of the conversation. It also lacked the humans knowing that it was a test, so they didn't properly "evaluate" it regardless.
Re:Fraudsters Used AI to Mimic CEO’s Voice
by _carl_jung on Tuesday September 03, 2019 @06:59AM
(#20865335)
No it isn't. The Turing test is a methodology to _try_ to "determine whether or not a computer is capable of thinking like a human being". It actually quite famously falls down at exactly this hurdle: mimicking the natural language patterns of humans[1].
Re:Fraudsters Used AI to Mimic CEO’s Voice
by Traster on Tuesday September 03, 2019 @05:53AM
(#20865057)
At best this attack used AI to create a text to speech engine whose output sounds the like the CEO in real-time. At worst it generated a number of pre-canned sentences that the scammers just replayed over the phone in the right order.
There's nothing in this scam that implies the AI is in any way responsible for the actual content of the conversation.
The source of that fact is probably a Wall Street Journal reporter, who heard it from the victim's insurer, who heard it from the police.
If they didn't catch the criminals, how is it possible for the police to find out they used "deepfake" or "AI"?
Whilst it's funny these guys did this with AI I can't help but feel that they could have done this with someone who is halfway decent at accents and saved themselves a lot of effort.
Imagine getting a call from your wife. Do you doubt right now if it’s her or something that just sounds similar.
This man could have know the ceo quite well.
Would you tell your wife your credit card number if she called you and told you her card is blocked for some reason and she needs to refill the gas tank or buy something?
So I asked who sold him that Music Man Petrucci 7-string in 2009(or so) and after he answered correctly, I followed through with his request.
My point being: it helps to have such a question prepared for everyone - or alternatively - a question concerning some fact about you that only people who know you could answer.
but the deepfake thing is a good story.
On the contrary, it seems like they did save themselves a lot of effort by deploying the de-centralised 'deepfake' AI model. It would be significantly hard for the opportunistic scammers to go out talent-scouting locally, trying to match the exact vocal characteristics of a person. It would also create a vulnerability, notably the person hired to impersonate the CEO.
Do you really think that most employees would need more than that?
But yes, at my current employer, I believe no one would do that just based on an email and a phone call. But then I dare say that faking Swiss German is an order of magnitude harder than faking English.
Someone has compared the typical cadence of Schwyzerdütsch to riding a donkey over very rough mountainous terrain, and I'm inclined to agree.
Also while Bitcoin community is creating new standards for collaboration, like multisig transactions and partially signed transaction format (which would help here in creating a bank-independent solution), I don't see the same standardization happening in the banking system.
The Turing test is "determining whether or not a computer is capable of thinking like a human being"
Not whether it can mimic the voice and speech patterns of one.
A computer capable of thinking like a human is "strong ai", or "general AI", or any of a few other poorly defined terms.
The turing test is a poor approximation of that goal. Computers can and will pass the turing test long before they're actually capable of real human thought.
Specifically, the turing test is merely having a human judge evaluate a free-form text-based conversation with a partner, and them being unable to distinguish a computer from a human in that scenario.
The main thing this instance lacks is the free-form nature of the conversation. It also lacked the humans knowing that it was a test, so they didn't properly "evaluate" it regardless.
[1] https://en.wikipedia.org/wiki/Turing_test#Weaknesses
There's nothing in this scam that implies the AI is in any way responsible for the actual content of the conversation.