For the attacker who wants the perfect botnet for their particular target but doesn't want to build it themselves, there is now Vawtrak -- a large botnet that can be broken down into smaller pieces and customized with a wide variety of web injects.
As described in a report released today by SophosLabs, Vawtrek is "apparently being used as part of a Crimeware-as-a-Service (CaaS) business model where the output of the botnet can be adjusted on demand, with financial data effectively being stolen to order."
Vawtrak (also known as NeverQuest and Snifula) is primarily after online bank accounts -- stealing credentials, sneaking around two-factor authentication, accessing accounts, transferring funds, and hiding the activity. It may disable anti-virus. It may install mobile malware. It's adaptable.
On Sophos's threat scoring model for botnet-based cybercrime infrastructure, Vawtrak is considered the single-most dangerous threat. It is delivered via exploit kits (usually Angler EK), spam emails claiming to be from financial organizations, or via loader malware like Pony Loader. Sometimes the droppers are issued as .exe files, but usually they're DLLs.
The real value of Vawtrak is specialization -- it targets banks in both Germany and Japan (and others), but approaches them differently. As the report explains:
The larger botnet can be broken down into smaller botnets that each have different web injects that are designed to target banks and other financial organisations in specific countries. The Vawtrak owners can then measure the success of specific campaigns and can divide the product of the whole botnet into subsections. New targets and campaigns can be set up as demand requires.
This allows for a commercial model where the Vawtrak operators create campaigns based on customer requests, selling the output of the botnet.
The operators offer a wide selection of web injects. For example, they can elicit the answers to the secret questions a user must answer to change a password. Or log a user out of their email to prevent them from seeing a legitimate message from their bank alerting them to a money transfer. Or add extra fields to Web forms, to collect extra info. Or coerce a user into entering a legitimate one-time password, then using it to carry out a fraudulent transaction.
Read the full report here.