Breaking download for breaking purposes

November 6, 2017 in Random ideas

This is a thought experiment only – it occurred to me that a number of available hacked sites that malware authors use for the purpose of payload delivery is making the the infection process not only easy and resilient. It also makes it easy to deliver a fragmented download mechanism that is similar to P2P technology and one that could potentially affect a number of security solutions.

Imagine having 2 or more hacked web sites – each of them hosting segments of the final deliverable e.g. it could be in a form of odd and even bytes, sectors, clusters, or the payload broken into pieces using some fancy, customized algorithm. The moment the malware loader (e.g. Office macro) is executed, it could download these pieces from various places and reassemble them on the host. Since they will be coming from different web sites, they will be different network streams associated with these downloads and as such every single security solution put in place will fail to detect an executable as a whole. This will therefore break the IDS, IPS, inline proxy, inline sandbox, the on-host EDR and AV, and whatever else is inspecting the network traffic, and in some cases – files.

There is also another angle to it; not having plain-vanilla .exe residing on hacked servers will make it harder to detect them – since they are just fragments of the final payload – no security solution will detect them as executables, hence they can escape .yara scans, etc. (file integrity monitors is a different story).

Obviously, one could simply encrypt the payload, or even encode it, but this is not the point here – many malicious samples still download pure vanilla .exes and they come from a single location.

Comments are closed.