Three Banks Lose Millions After Wire Transfer Switches Hacked
mask.of.sanity writes "Criminals have stolen millions from three unnamed U.S. banks by launching slow and stealthy denial of service attacks as a distraction before attacking wire payment switches. The switches manage and execute wire transfers and could have coughed up much more cash should the attackers have pressed on. RSA researcher Limor Kessem said, 'The service portal is down, the bank is losing money and reliability, and the security team is juggling the priorities of what to fix first. That's when the switch attack – which is very rare because those systems are not easily compromised [and require] high-privilege level in a more advanced persistent threat style case – takes place.'"
I must be missing something -- did these people transfer it to an account then go withdraw millions in cash quickly? Or did it take months for it to be discovered?
I can't conceive of any other way that would insulate against a reversal, no matter how many accounts and banks around the world they forwarded it to. Even Swiss banks go along with obvious criminality investigations nowadays.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
slow and stealthy denial of service attacks
I don't think a DOS can be stealthy......if it's denying service, are people going to notice?
A stealthy DOS is when the attack looks like a normal occurrence, and not an attack. It is not the DOS that is stealthy, it is the attack or, rather, the reason for the lack of service.
It is a very neat thing, actually. Say you have a very long, segmented fence. There are 1000000 segments, and every day 1 of those will break and stay broken for 10 seconds. You can't explore that, because it is random, and you can't try all 1000000 segments in 10 seconds. However, if you can force the dice and make a specific segment tail, you can be there and exploit it, because you know which one and when. To the external observer, however, it was just a normal, run of the mill segment fail.
It is the same concept. The failure is there, they notice it, but it is done in such a way they don't notice it is an attack.
morcego
You can put authorization codes in transactions, but if they aren't digitally signed, you can alter them in transit. Maybe banks should start exchanging signing keys and not transfer authorization codes?
I was promised a flying car. Where is my flying car?
The bottom line is that we need to harden up our defences more and more. We may even have to disconnect essential financial infrastructure from the internet and bring it back onto a completely private network that it costs a substantial amount of money to join and be authenticated to. It should come with the proviso that any device connecting to it, could also not be connected to the internet or an unknown intranet device at the same time. This would not be bulletproof, but it would substantially reduce the risk.
Meus subcriptio est nocens Latin quoniam bardus populus reputo is sanus callidus
You would be amazed - or maybe shocked - to see some of the banking systems out there. I have worked for several financial institutions and their systems are usually very very old legacy crap stuck together with bubble gum and faith. One place was dealing with 70% of the countries financial messaging and they were not using transactions, if there was a problem (and there often was) messages were lost. Asked if I could change it to use transactions, couple lines here, couple lines there.
NO.
Why?
Cost to test would involve the entire country and would cost millions.
OK.
So they are still losing messages.
There are three kinds of falsehood: the first is a 'fib,' the second is a downright lie, and the third is statistics.
If the banks had a way to extract more money from us, wouldn't they already be doing it? Why would they wait until they were hacked and lost money to raise prices, if they thought it would increase their income?
Give me Classic Slashdot or give me death!