How deepfakes scammed clerk for 26$ million in Zoom
HTML-код
- Опубликовано: 29 июн 2024
- 👾 Identity Fraud Report: sumsub.com/guides-reports/ide...
Our new video uncovering the investigation into the largest corporate fraud involving deepfake technology. Learn how a Hong Kong office worker transferred $26 million to local accounts after being tricked by what he believed were senior colleagues.
We've analyzed available sources to reconstruct the attack scenario. Join us to understand:
- The selection process of the victim
- The organization of the attack
- Reasons behind the victim's compliance
In addition, we'll discuss strategies for safeguarding against realistic deepfake attacks. Remember, it's not just celebrities or corporations at risk-fraudsters are exploiting the identities of ordinary people too.
Timecodes
00:00 - Intro
02:16 - Phase 1. First contact
04:00 - About the letter
04:57 - Human error
10:18 - Phase 2: Preparing
14:05 - Patient Zero
21:08 - Phase 3. Call
24:50 - Conclusion
#deepfake #AIfraud #deepfakefraud #sumsub
Sumsub - empowering compliance and anti-fraud teams to fight money laundering, terrorist financing, and online fraud.
More about us:
sumsub.com
/ sumsubcom
/ sumsubcom
/ admin 
Наука
Found this video / channel through shorts! Thought you’d like to know 😊 great video
Yeee. I've been following their stuff for maybe over a year now, or close to it, I guess. Welcome 🤝. I do some ethical hacking, bug bounties. It's not very easy usually, and I've gotten nothing so far. 😅 Shalom.
Thanks, mate! So great to have you with us, welcome!
Same
@@ReligionAndMaterialismDebunkedtoday I learned there’s a such thing as ethical hacking. Huh, that’s interesting!
Same!!💙
Easy way to protect yourself from impersonation: keep your thick regional accent and local slang.
Makes it much harder for AI to convincingly emulate your diction/lexicon.
Maybe we could see a wider cultural shift towards a less corporate-talk style of communication?
Something tells me that an accent isn’t too difficult to learn for even today’s AI. That’s not the point of this scam, however. They even showed examples of how goofy some of those AI’s were.
My regional accent is approximate received pronunciation. Wtf am I supposed to do?
@@Bellicosylearn a new one
AI would adapt in time, though.
{:o:O:}
@@ansfridaeyowulfsdottir8095 it would, and at a faster pace than humans perhaps. Scary! What if ai started to dictate the local accents that humans copied and used?
Glad that in Norway we get tons of personal questions just to do anything. Like my bank ask me so much stuff only I whould know its annoying, can’t get anywhere if you don’t answer
Very interesting & informative video!!
However, I think you really should provide links/URLs to the research articles/papers you show snapshots of and/or cite/reference.
Awesome video, you got my sub! I love stuff like this, its very interesting. Something i try and teach my staff is that vulnerabilities will always occur as long as the human component exists. At some point, even the most alert may fall for a phishing attempt. Someone, is always expecting something: an email, a delivery, an invoice, confirmation code, SOMETHING. If a hacker is good at social engineerings and OSINT, they will find and time these events.
Awesome, thank you!
Wouldn't it be easier to just lock the account information behind the customer's security questions? That way an employee can never break policy because they are physically unable to access that information without authentication.
In some cases, this already exists. However, depending on the scope of said employee, they might be required to check customer accounts. Some companies instill certain clearance levels for need-to-know if your job requires you to have access to someone's account. Example: if your job is to monitor fraudulent activity on customer financial accounts, you would need to have access to those accounts. However, if your job is to look at complaints for example, you shouldn't have access to customer's financial accounts and wouldn't be given this clearance.
Unfortunately, the scammers must already know this because they use the CEO or high level managememt to get someone to do the work, which means that the ones who do have access are the ones targeted :(
There is a very good historical example from Germany, Friedrich Wilhelm Voigt, better known as "Hauptmann of Köpenick" He put on a uniform and confiscated a fortune
Glad to be a subscriber of your channel ❤
Yes!! You haven't uploaded a video like this in a while. Glad to see material like this being uploaded again ❤
Thanks a lot!
We need deep fake detection tools built in to conferencing, email and media consumption (such as YT) platforms. Thats minimum for now. Ultimately, as fakes improve, that may not be enough.
Wow thanks for this content.
Having to do 2-step verification when you are Zooming with your colleague is crazy
@Sumsub !! Very shocking!! Great video 👍
Countries world wide need to get serious about AI and how it can and can’t be used and create severe penalties for crimes committed using AI.
Just make a video for Dark net Market ?
Amazing content! Keep up the amazing work! We love to know how this stuff works!
Thank you! We will!
Great work
Thank you so much
Great Tutorial
Glad you think so!
第一次聽到呢個新聞，咁嚇人嘅😮
Easy to avoid not using machines
Two keywords to remember: quick and secret
This Channel would do a real service to fan bois! of all ages and genders on what is and is not AI because right now, it is an overused term much like any other Marketing term.
26:00 Let's make it mandatory that any face mask must have a clear label, even if you make the mask yourself. Finally a law that'll avoid especially the high profile OSINT attacks! 🙃
No. That law is barely helpful to detect the easiest fakes online. People need to adapt. Banks, CEO, government officials, your mom, you and I need to adapt as the world changes and we need to help each other.
Don't mind me, just being pedantic.
$26 mill*
I'll be on my way ᕕ( ᐛ ) ᕗ
Badly scripted video, digressing here and there
i like been the 20 comment on this channel, keep it up !!!
This has been oddly inspiring.