Typically, banking customers discovered they'd been victimized by cyber-crime when they discovered their bank accounts emptied. No more. According to this report, online thieves are getting craftier at covering their tracks to go undetected for longer stretches of time.In its Cybercrime Intelligence Report [.pdf], anti-virus provider Finjan details how cyber-criminals are using Trojans, such as the URLzone bank Trojan, and new anti-fraud detection tactics to sidestep quick identification of the crime by both banking customers and the banks themselves. As has been the case for some time when it comes to building or using malware, the underground market is alive and robust. According to this report, cyber-criminals can buy an online crime toolkit, such as LuckySploit, for anywhere from $100 to $300. These kits contain exploit code and a management interface for injecting attack software into Web sites. These don't always have to be malicious Web sites, mind you, they can also be legitimate Web sites with vulnerabilities that leave them susceptible to attack. Unfortunately, many are.

But, unlike the typical attack on a bank account, the thieves Finjan studied (in this case using the URLzone Trojan) to attack German banking customers actually hid the fraudulent transactions from their banking customers. They did this by manipulating the account statement information that is displayed on the customer's screen. "In order not to raise any suspicious[sic] by the user the malware verifies that the user will see only what he is [sic] expect," wrote Daniel Chechik in a Finjan's blog that went into meticulous detail on how this malware operates.

The malware also, through code manipulation, tricks end users into accepting fraudulent transactions. This technique also enables URLZone to bypass the security offered by a One Time Password (a new password that is generated every time a user accesses their account, which is supposed to provide more security than the more common static password).

Now, with the banking customer kept ignorant regarding the fraudulent transactions -- it's time to make certain that no anti-fraud alarms go off at the bank itself. The crooks try to active this by, according to the report, by making sure the victim's balance stays positive, the fraudulent transaction is not too high, and the amounts are random.

Let's take a step back for a moment, to make sure we appreciate the complexity of this attack: First, the crooks managed to either infect a legitimate Web site, or setup a fraudulent site to lure victims. Second, they must have presumably managed to sidestep any anti-virus that was installed (which we recently revealed not to be very reliable, as it turned out). Third, bypass one-time-password authentication by making certain that the attacker has commandeered the browser session during a banking transaction. They then fake the statement presented to the consumer online. And, finally, craft transactions so that they avoid detection by the anti-fraud controls the bank may have in place.

That sophistication reveals the level of motivation, and skill, banks and consumers are up against.

Is all of this effort worth it?

If Finjan's math is correct: yes.

The thieves managed to haul about €12,000 a day.

For security and technology observations throughout the day, follow me on Twitter, @georgevhulme.