IBM research has discovered a concerning development in some new Ice IX configurations that are targeting online banking customers in the U.K. and U.S. Ice IX is a modified variant of the Zeus financial malware platform.

In addition to stealing bank account data, these configurations are capturing information on victims’ telephone accounts. This allows attackers to divert calls from the bank that were intended for its customers to attacker-controlled phone numbers. The fraudsters are presumably executing fraudulent transactions using the stolen credentials and redirecting the bank’s post-transaction verification phone calls to professional criminal caller services that approve the transactions.

How Ice IX Logs In

In one attack captured by IBM researchers, at the point of login the malware steals the victim’s user ID and password as well as the answers to the user’s memorable information/secret question, date of birth and account balance.

Next, victims are asked to update their recorded phone numbers (home, mobile and work) and select the name of their service provider from a drop-down list. In this particular attack, the three most popular phone service providers in the U.K. are presented: British Telecommunications (BT), TalkTalk and Sky. Here’s an example of the webinjection the user sees in his or her browser:

To enable the attacker to modify the victim’s phone service settings, the malware then asks the victim to submit his or her telephone account number. This is very private and sensitive data — typically only known to the phone subscriber and the phone company — and is used by the phone company to verify the identity of the subscriber and authorize sensitive account modifications such as call forwarding. The fraudsters justify this request by stating that the information is required as part of a verification process caused by “a malfunction of the bank’s anti-fraud system with its landline phone service provider.”

Here are the webinject messages presented to BT, Sky and TalkTalk users:

Fraudsters are increasingly turning to these post-transaction attack methods to hide fraudulent activities from the victim and block email and phone communication from the bank. This allows attackers to circumvent security mechanisms that look for anomalies once transactions have already been executed by the user.

Deterministic detection security mechanisms, like IBM Security Trusteer Rapport, search for specific malware crime logic footprints before transactions are submitted and allow the online banking application to stop fraud by changing business flows (block money transfers, decline/add payee, limit amounts, etc.). Uniquely, they are not vulnerable to post-transaction attacks.

Thanks to Tanya Shafir from the IBM analysis team for providing the intelligence for this post.