Zeus botnet raid on UK bank accounts under the spotlight
More details of sophisticated cyber-blag emerge
Analysis More details have emerged of how security researchers tracked down a Zeus-based botnet that raided more than $1m from 3,000 compromised UK online banking accounts.
Bradley Anstis, vice president of technical strategy for M86 Security which discovered the attack, said hackers began the assault by loading compromised third-party sites with a battery of exploits designed to infect visiting PCs with variants of the Zeus banking Trojan.
Phase one of the attack used the Eleonore Exploit Kit and the Phoenix Exploit Kit to load Zeus onto compromised machines through a battery of browser and application-based vulnerabilities and drive-by download attacks. The main attack revolved around the use of version 3 of Zeus to steal money from online bank accounts.
Version 3 of Zeus is focused on stealing the login credentials of online bank accounts. Older versions also snaffled user names and passwords for social networking sites and online services as well as online banking credentials.
Version 3 also differs from its predecessors in the use of encrypted tunnel (HTTPS) in communicating back to the command and control servers.
The use of a different strain of Zeus means the M86 researchers are sure the attack is unrelated to an otherwise similar attack involving 100,000 compromised UK bank accounts that was the subject of an alert by transaction security firm Trusteer last week.
After noticing a pattern of possible attack, M86 researchers deliberately infected a machine in order to identify a command and control server associated with the botnet which was hosted in Moldova. They then used exploits to break into the poorly-secured system where they found logs recording the activity of compromised bank accounts.
The logs also revealed that 3,000 online banking accounts had been victimised between 5 July and 4 August alone.
Both business and consumers accounts at one single unnamed UK bank were hit. Anstis declined to say what bank was involved beyond saying it offered free internet security software to its clients. Those hit had therefore either not taken up this offer or were running defences that failed to block malware from infecting their systems.
M86 said it was in touch with the bank involved in the attack.
The Zeus Trojan associated with the account ran a "man in the browser" attack. While earlier versions simply recorded banking login credentials and forwarded them back to hackers, the latest version of Zeus also performs illegal online banking transactions. The malware is sophisticated enough to wait for the entry of secondary authorisation data needed to make transfers - such as a date of birth - by victims onto compromised machines.
Accounts balances from compromised machines were manipulated to disguise transfer from compromised accounts to phishing mules, located in the UK. These transfers were only carried out when the account balance was over £800, a tactic possibly designed to avoid the early detection of fund transfers from compromised accounts by making sure marks still had enough money to obtain cash from ATMs.
Anstis said the attack was ongoing. M86 has passed on its dossier on the attack to law enforcement agencies, which Anstis refused to name.
For UK-targeted attacks involving malware the Police Central eCrime Unit at Scotland Yard would be the natural point of call, though it proved impossible to confirm any investigation by the Met at the time of writing.
More details of M86's investigation and the malware behind the attack can be found in a whitepaper here. ®
This article was updated to remove erroneous information concerning Mac OS X. The OS is not vulnerable to the crimeware kit.
Add another layer of security.
Isn't it standard proactice to have the RSA security tokens(or other compnay) used by most IT companies. Why can't banks give these to customers. The loss cost to banks is probably going to be greater than the cost of the fobs somewhere down the line. I can't see the banks taking it in the ASS like this forever. They will try and offload responsibility to the user, just like they did with Chip and Pin.
Why cant we have more malware?
mac users are safe from this windows software.
Yes it WOULD stop a man in the middle attack
...because you enter more information into the device than the PIN, so the token is dependent on the specific operation you are performing, as I said in my earlier reply.
The man in the middle can therefore only repeat the same transaction you have just done, not generate a different transaction (such as transferring money into his account).
Presumably you took what the original poster said literally. But Nationwide's card reader doesn't generate a token from just the PIN, and I bet Barclays' doesn't, either.