Monthly Archives: February 2014

Researchers crack Bitcrypt ransomware

There are 256 (28) different bytes and only ten different digits. So if your secret (RSA) key consists of 128 digits rather than of 128 bytes, the entropy of the key (that is, the amount of ‘surprise’ to an attacker) is a whole lot lower.

No shit, Sherlock. Apparently, this somewhat basic fact was beyond the understanding of those who wrote the Bitcrypt ransomware, probably inspired by the sad success story of CryptoLocker. In practise, it meant the difference between “only the NSA can crack your key” and “anyone can crack your key”. Two researchers from Airbus cracked the key and thus were able to restore the encrypted files on a friend’s computer, without paying the 0.4BTC ransom.

More at Virus Bulletin here.

Windows Error Reporting used to discover new attacks

Security firm Websense published a report that explains how they can use error reports generated by Windows to discover new targeted attacks (‘APTs‘ in security hipster speak). It’s interesting, but it barely touches on the fact that these reports being sent in cleartext is also a serious problem. I wrote a blog on both sides of this issue for Virus Bulletin.

(I’m not sure if anyone is reading my musings here, but I thought it might be a nice idea to link to things I write elsewhere. I also hope to find inspiration to write the odd thing that has nothing to do with computers or security at all.)

The Return of Qakbot

Together with João Gouveia of AnubisNetworks, and using their real-time feeds, I’ve been looking at Qakbot, a piece of malware that was huge in 2011 and had since disappeared off the radar.

We found that Qakbot is still active and there are at least 20,000 infected devices. The command and control protocol has progressed from version 2 back in 2011 to version 8 today. We cracked the obfuscation used in earlier protocols, but are still struggling with version 8, which appears to use encryption rather than obfuscation.

I tried a large number of obvious and slightly less tricks to crack the protocol (including RC4, which I didn’t mention in the blog post), but so far to no avail. If anyone has any suggestions on how the encryption might work, we are of course happy to learn of it.

Still, I am quite content with the research we did, which will hopefully contribute to the knowledge of and the fight against Qakbot. The blog post is here. (NB the original blog post is not available any longer; an archived version can be found here.)