Saturday, August 3, 2019

The Application Era of Digital Forensics

I started in digital forensics in 2002 which seems like a lifetime ago when I look at the digital forensics world today compared to back then.  One of the primary changes is that the center of gravity for host-based digital forensic examination has moved from file systems to applications.  There has been quite a bit of discussion about this trend over the years so it’s not a unique observation on my part, but I’ve certainly watched in wonder as digital forensics methods and tools have changed to adapt to this reality.

Around the turn of the century, the primary focuses of host-based digital forensic examinations tended to be web browser artifacts, document and photo metadata, file system artifacts, operating system log files, and email parsing.  Forensic examinations weren’t exclusively focused on these areas, but most digital forensics exams would involve at least some of them as their core components.  If you had proficiency in these areas, there was an excellent chance that you were a competent digital forensics examiner.   In my mind, the key differentiator at that time was whether an examiner understood file systems well enough where they could understand and articulate concepts such as how date and file data manifested itself inside of a master file table.  If I interviewed someone for a job and they told me that there were only four date stamps inside of NTFS master file table record, I knew they were likely just someone who drove a digital forensics tool and didn’t really understand much more beyond that.

As I look back on it now, the relatively lack of diversity in digital forensic tools at that time makes more sense to me. It was very common for a digital forensics shop to use EnCase as their primary digital forensic tools along with FTK as their indexing/email forensics tool.  FTK was also nice to have around for a second opinion on what EnCase was telling you in regards to the file system. Both tools did web parsing reasonably well, but many of us used the very fine NetAnalysis tool for web browser history forensics.  We didn’t have a vast diversity of tools because there just wasn’t the business case for them. Our existing tools generally did what we needed them to do.

If someone were writing the history of digital forensics, the advent of Magnet Forensics Internet Evidence Finder (IEF) would likely show up at the start of a chapter talking about the switch of focus from file systems to applications.  IEF became very popular, very quickly because it was designed with a focus on parsing application information whether that application was part of the operating system such as a native web browser or whether it was a third-party application such as a third-party web browser, chat program, email client, P2P client, and the like.  Of course, it also did a great job parsing a whole host of operating system artifacts just like the other tools did, but the long-term secret sauce of IEF was that it was focused on application artifacts in an era when apps were becoming the primary focus of consumer technology use and spending.

IEF was a great product and, as it turns out, the sign of a great business strategy in the making.  The computing world was moving from one where user activity was focused on things such as web browsers and office applications to one where users were using an amazing diversity of applications primarily on their mobile devices.  It’s not that we gave up using web browsers and office applications, but they were just part of the greater mix of applications being used.  Just take the Apple store as an example.  It opened up in 2008 with about 500 applications available on it and by 2017 it had over two million applications.  

This rise in applications drove changes in the development of digital forensic tools. The companies that have focused their development work on applications have generally done very well for themselves.  Magnet is an obvious example since it went from a one-person shop to a global digital forensics company by riding this wave.  Other examples are MSAB, Cellebrite, and Oxygen Forensics who have done very well for themselves by also capitalizing on this trend and creating products to address it. 

It’s not that the core digital forensics skills such as file system forensic analysis are obsolete. Far from it.  You still have to load yourself up on materials such as the magnificent Harlan Carvey books especially if you are investigating network intrusion cases in an enterprise computing environment, but now you also have to understand how operate as a digital forensics examiner in this application-centric mobile device era.

What do I mean by this?  We’re in this new era of digital forensics where examiners are going to have to get comfortable with being even less able to rely on their forensic tools for support than before.  You can’t rely on your commercial tools any more than you did in the past file system focused era.  In that era, the name of the game was going beyond your tools and understanding the underlying technology well enough so that you could validate the output from your tools.  That’s still the case now.  In this new era, the commercial forensic tool developers will use their finite development capacity to support only the applications that are broadly used and whose support will drive people to purchase their tools. Because of this, the digital forensics community will need to provide their own support for applications that aren’t covered by commercial digital forensic tools. 

What will this support look like? It’s going to come in the form of enterprising digital forensics people using their knowledge of parsing technologies such as JSON and SQLlite to create their own tools and scripts to investigate these artifacts.  In some cases, this will take the form of creating scripts that leverage existing digital forensic tools or creating tools/scripts that work as standalone solutions.  As always, we’ll still need to have the ability to double check what these tools are telling us so that we can validate the results. I use the term “have the ability” because I understand that an individual examiner can’t be expected to know everything such as having the ability to comfortably parse a particular Python script or JSON artifact.  The ability to do these things might take the form of being able to leverage someone you know to do this work for you.  Digital forensics is a team sport and one of the most important tools in your inventory is a list of people who are willing to help you out when you are in a bind.

Cryptocurrency wallets are a good example of all of this.  Cryptocurrencies are the primary payment system of the online underground economy and wallets are applications that allow people to interact with cryptocurrency blockchains so they can send and receive transactions. There are just too many wallets for the digital forensics companies to provide support for all of them and it’s going to take the community creating tools to parse them to provide the necessary support for cryptocurrency related examinations.  This is not an original point on my part since Jad Siliba made this observation about cryptocurrency forensics at the 2019 Magnet User Conference (MUS2019) this year and those comments have stuck with me ever since. 

Various and Sundry

I’m still trying to get back into a monthly blogging tempo and I have another AFoD interview in the works as I write this.  I’m coming towards the end of a very heavy conference presentation schedule this year that started about the time of MUS2019 and will end for me at the end of the upcoming Dallas Crimes Against Children Conference.  Thanks to everyone who came out so see me speak on Business Email Compromise and Virtual Currency Investigations over the past four months or so.  I’ve enjoyed getting to meet so many new people as well as to finally meet some people who I had only known electronically through the years.   I’ll still be doing some presentations in 2019, but I think the heavy digital forensics conference season is pretty much done for me and most of the people I know who do speaking on this circuit.  I’m going to be working on developing some new talks for the 2020 season.

I basically wrote this blog post in my head while at the MUS2019 conference after listening to people like Jad and attending some great talks.  For example, Alex Brigoni did a magnificent talk called “Unsupported Apps. What Can Be Done? A Methodological Approach to Mobile App Forensics” that covered must how digital forensics people should be approaching this new application-centric era.  He’s a razor sharp fellow and you can find him over at his blog and on the Twitters.  You should also read an interview he did on the Magnet blog regarding application forensics.