Encrypted and hidden files put eDiscovery at risk


There are some pretty nice freeware applications available which allow a user to encrypt and hide files/data/electronic records in plain sight on their computers. Can this pose a problem for IT and corporate legal?  Let me put it this way…how would you find and place ESI that’s encrypted or is both encrypted and made to look like something else on a litigation hold?

Does the fact that encryption applications present in a corporate infrastructure make claims of spoliation if the files can’t be found or decrypted more likely? Is this a problem you should even worry about?

It’s a stretch but in some circumstances this capability could significantly raise your eDiscovery risk. To illustrate this problem further I will specifically talk about an application called TrueCrypt which is a free open-source disk encryption software application for Windows 7/Vista/XP, Mac OS X, and Linux.

TrueCrypt is an application for creating and maintaining an on-the-fly-encrypted volume (data storage device as opposed to a single file).This means that you can create an encrypted volume capable of storing many encrypted files which to casual observers, looks like a single file. On-the-fly encryption means that data is automatically encrypted or decrypted right before is loaded or saved, without any user intervention. No data stored on an encrypted volume can be read (decrypted) without using the correct password or correct encryption keys. There are several encryption algorithms available in the application but the most secure is the AES 256-bit key algorithm, the same one used by the federal government in many instances.

Files can be copied to and from a mounted TrueCrypt volume just like they are copied to/from any normal storage device (for example, by simple drag-and-drop operations). Files are automatically decrypted on-the-fly (in memory/RAM) while they are being read or copied from an encrypted TrueCrypt volume.  Similarly, files that are being written or copied to the TrueCrypt volume are automatically being encrypted on-the-fly (right before they are written to the disk) in RAM.

Now, to make matters worse (or better depending) TrueCrypt also can create a hidden encrypted volume within the visible encrypted volume.

The layout of a standard TrueCrypt volume before and after a hidden volume was created within it. (Graphic from the TrueCrypt manual)

The principle is that a TrueCrypt volume is created within another TrueCrypt volume. Even when the outer volume is mounted and visible, it would be impossible to prove there is a hidden volume within it or not, because free space on any TrueCrypt volume is always filled with random data when the volume is created and no part of the (dismounted) hidden volume can be distinguished from random data. Note that TrueCrypt does not modify the file system (information about free space, etc.) within the outer volume in any way.

So to put it another way, an employee trying to hide data from a discovery search could first create an encrypted volume on their hard disk or some other removable device such as a USB stick and store encrypted data on it. Even more diabolical, they could move some innocuous data to it as a decoy and store the real data on the hidden volume inside the original volume. This capability provides the employee plausible deniability in the case of corporate legal or IT forces the employee to decryption the volume they can see.

So the big question is this; how would you as a discovery auditor even know of or find these hidden and encrypted data volumes? In reality it’s not easy. You have to go into it looking for hidden and encrypted data. There are some forensics applications that will at least find and flag encrypted files and volumes including the TrueCrypt format. I am unable to determine if these forensics applications can find and flag hidden volumes.

As a test, I setup a 10 GB TrueCrypt encrypted volume on this computer and named it “Attorney Communication” in a folder I named “contracts”. To the casual observer all they see is a large file in a folder called “contacts” (see below).

Within that encrypted “Attorney Communication” file I copied four decoy files to make it look like those were the important files I was keeping encrypted just in case I am forced to open the encrypted volume by legal (see below).

As you can see above, you can’t tell by looking at it that it contains the hidden 8 GB volume I had also created. That hidden volume is accessible only by typing a totally different password.

The hidden 8 GB TrueCrypt volume on this computer

So how do you find these hidden volumes and files if the employee is not cooperating? If you suspect the employee has been using this technology the first obvious step would be to do a search of the employee’s hard disk looking for an application called “TrueCrypt”. This would be a dead give-away that the employee could have encrypted and hidden data volumes on their computer. This is not  certain because the employee could have installed the TrueCrypt application on a USB stick, which does not integrate with Windows, so when not plugged in to the computer, there would be no trace of the TrueCrypt application.

A second way to find potentially encrypted volumes would be to search for very large files. Usually encrypted volumes will be larger than normal files because they are just that, a large space to store many files. So you could do a Windows search for files over 10 MB and see what you get. An indication would be a large file with no applications associated with it. By this I mean that when you double click the file the system doesn’t recognize it as a standard Windows application and displays the “Open with” dialog box shown below:

That leaves the problem of discovering the hidden volume. A sure but very slow process to test this possibility would be to copy a bunch of files into the encrypted volume, if the employee has opened it, to see if the available storage space id equal to the volume size.  For example the file properties in Windows states my encrypted volume is 10 GB in size but in this example the employee only has 5 MB of files stored in it. To test to see if the volume contains a hidden volume, you could copy an additional 9.95 GB of data into it to see if you get a “volume full” message before all of the data was copied into it. If you could only copy an additional say 1.95 GB before the “volume full” message was received, that would indicate that a hidden 8 GB volume exists.

A faster way to get an indication of hidden volumes is to use a large file finder tool. I found one called “Largefiles3” which had a surprising capability. In this case I ran the application looking for files larger than 10 MB on the C drive.

The interesting capability here is that it found the encrypted volume I named “Attorney Communication” but it determined its size to be 2.147 GB not the 10 GB shown in the Windows file system data. This is because I had created an 8 GB hidden volume inside the “Attorney Communication” volume thereby only leaving 2 GB for the original volume. This is a huge red flag if you are looking for it. Now, without the password you still can’t access the original encrypted volume or the hidden volume but at least you would know it exists and can apply pressure to the employee.

So how do you prevent these encryption applications from putting your eDiscovery processes at risk? The most obvious one is to include in your employee computer use policy a statement prohibiting the use of these types of applications with stated punishments if not followed. This will stop general employees from using this kind of application but will not deter those employees bent on breaking your rules. The obvious next step is to sample and audit your employees to see if these applications are being used. For corporate legal, the main thing you want to establish is your “good faith intent” to make sure your eDiscovery processes are defensible.

Advertisements

Cloud Storage is not as secure as local storage…Really?


A December 13th posting in the Harvard Business Review Blog titled; Cloud Computing? Not So Fast — Unintended Consequences of Recent Disclosures, seemed to conclude cloud computing/cloud storage was inherently less secure than traditional local or on-premise storage of sensitive records.  The blog entry tried to site a couple recent cases to prove the point. One example used was the wikileaks revelations.  The implication was this massive amount of leaked information was (more easily) leaked because it was “stored in the cloud” verses somewhere else…such as an organization’s NAS or SAN.

The implication of the blog posting is inaccurate in that it assumes the top secret federal government controlled files were stored in a third parties data center with little or no security and this was why army private Bradley Manning was able to easily steal a such huge cache of files/records.

I don’t know for a fact, but my guess would be that this data was held within the federal government infrastructure and was not being stored at a third party data center. But even if the data was held at a third party facility, security of the data is key.

There are several definitions of cloud storage but the one most referenced is; cloud storage is storage accessed over a network (internal or external) via Web Services APIs. To many in the business world, cloud storage is a service which allows an organization to store electronic files/records in a third party remote location. Organizations will look at this service for a couple of reasons; first it may be less expensive then purchasing and maintaining their own local storage resources. The second reason is for increased security; sensitive data can be better protected from employee leaks etc.

The miss-used wikileaks example aside, it may be the author doesn’t fully understand the technology involved in on-premise storage verses cloud storage. In my experience, most organizations don’t protect the data their employees store locally with encryption or other safeguards. Many of the cloud storage providers encrypt data being stored to a cloud repository as it arrives at the storage facility. A couple providers encrypt it before it leaves the customer’s location; and of course the encryption key is known only to the owner of the data not the storage provider.

Top tier cloud storage providers store their customer’s encrypted data in class 4/5 secure underground data centers. Many Fortune 100 companies believe this type of storage is higher quality than letting their employees store sensitive data locally…