Citadel Trojan Targets VPN Credentials at Airport
Researchers at Trusteer recently uncovered a variant of the Citadel Trojan targeting the virtual private network (VPN) credentials used by employees at a major airport.
The firm would not disclose the name of the airport because the situation is being investigated by law enforcement. However, the airport is an international hub that services more than 30 million people annually, Oren Kedem, director of product marketing at Trusteer, told eWEEK.
"[The airport] took their employee VPN access down immediately last week, and it has remained down since," said Kedem. "They have told us they are changing their authentication mechanism immediately and are also looking into technologies to protect endpoint devices â¦ [such as] unmanaged laptops and desktops that access the VPN site."
The Citadel Trojan is built on the source code of the infamous Zeus Trojan, and is typically linked to the theft of banking credentials. In July, researchers at RSA's FraudAction Research Labs said Citadel's developers had announced plans to pull the software off the underground market and only make it available to existing customers and a select group of new ones. As of July, the Trojan sold for as much as $2,500, with additional plug-ins going for an average of $1,000 each.
In this case, the Trojan is being used in a sophisticated man-in-the-browser attack that is targeting the VPN credentials of airport employees to gain access to credentials that can be used to access internal airport applications, Trusteer reported. It is unclear how the victims were initially infected, though Kedem speculated it could be through spear-phishing attacks or drive-by downloads.
"Virtual private networks are often used to secure Internet connections between a business' private network and a remote employee," Amit Klein, CTO of Trusteer, explained in a blog post.
"This attack," Klein added, "uses a combination of form-grabbing and screen-capture technologies to steal the victim's username, password and the one-time passcode generated by a strong-authentication product (we have also contacted this vendor)."
The first phase of the attack uses form-grabbing to steal the username and password entered into the log-in screen. The second phase relies on screen capture to take a snapshot of the image presented to the victim by the strong-authentication product that offers both single-channel and dual-channel authentication options. The company declined to identify the strong-authentication product by name.
"The dual-channel mode sends users a one-time PIN via [Short Message Service] SMS or a mobile application," Klein explained. "The single-channel mode is activated when users select the 'Get Image' option at log-in. This prompts the strong authentication product to generate an on-screen CAPTCHA of 10 digits. The user then maps their original (static) password to the string of digits in the image to produce the "one-time code."
"This security measure prevents the form grabber from capturing the actual static password," said Klein. "This is where the screen-capturing feature in Citadel kicks in. By capturing the image, the attacker uses the permutation of digits, along with the one-time code stolen by the form grabber, to reproduce the static password."
The situation underscores that enterprises that rely on strong authentication products can still be compromised, said Klein.
"This is especially true in the case of unmanaged or [bring-your-own-device] BYOD endpoints," Klein blogged. "Since these devices are exposed to threats that would otherwise be filtered at the enterprise perimeter, they are much more vulnerable to infection from advance malware like Citadel, Zeus, SpyEye, etc. When an infected unmanaged device connects to the enterprise via a VPN or some other remote access method, malware can circumvent strong authentication systems â¦ [and] this attack is especially troubling, given the potential impact on air travel security and border control."