Who Built the 'Taj Mahal'?

A new piece of sophisticated malware has been made publicby researchers at Kaspersky. While the complexity and the sophistication of the malware suggest that it is the product of a nation-state, it bears none of the code fingerprints of any known nation-state hacker group.

Indeed, there are no code similarities between that which Kaspersky found -- named Taj Mahal -- with other known malware of any origin. That the entire code base used in the malware is novel is a big deal. One of the ways that analysts figure out where an attack originates is by comparing the code to that which has been seen before. With nothing to compare the code to, Taj Mahal keeps its origins secret.

The malware was found by Kaspersky in late 2018 to be targeting a single Central Asian diplomatic agency, which it will not name. It has two parts, or packages, named "Tokyo" and "Yokohama," which together contain 80 malicious plug-in modules. This is an extremely high number of modules, and the most that Kaspersky has ever seen in an APT tool.

Tokyo is the initial breaching malware and functions as a backdoor. It uses the standard PowerShell tool when it connects to a command-and-control server, and then plants the Yokohama payload spyware.

The sophistication of the malware is evident when file exfiltration is observed. If a USB drive is inserted into an victim's machine, TJ will scan the drive contents and uploads a list of them to the command-and-control server. The threat actors can then decide which files they want to exfiltrate.

But there is some really slick follow-up going on here. Say that the USB drive has been taken out of the machine by the time the actors have made up their mind. The malware can monitor the USB port for the same drive to exfiltrate that file, and send it to the command server the next time it appears. That is some fancy code work going on, making the malware even more dangerous.

The spyware has other modules that allow it to flag files that have been burned to a CD, or put into a printer queue. It can also take screenshots when recording VoiceIP app audio, as well as stealing Internet Explorer, Netscape Navigator, FireFox and RealNetworks cookies.

The stealth aspects of this malware are quite remarkable. The new code that is used is obvious (if it's never been seen before then you can't match signatures on it), as well as the behavioral detection avoidance. Target files are named randomly between reboots, as an another example. It helped the malware avoid detection from August 2013 until April 2018, based on the compile times found in some of the modules.

Kaspersky has published the malware's Indicators of Compromise.

Now that its cover has been blown, the security community may expect whoever is behind this will mount development of a new (and unknown) code base with the same purpose in mind.

— Larry Loeb has written for many of the last century's major "dead tree" computer magazines, having been, among other things, a consulting editor for BYTE magazine and senior editor for the launch of WebWeek.