July 20, 2011
For the past five years it seems like almost everything in information security has focused on application security and, for the NetSPI consulting practices, our application security business (app pen testing, code review, etc.) has significantly increased. In that time, we have seen areas like network and systems vulnerability assessments change due to the commoditization of those services. Qualys, nCircle, and Rapid7 have all created a less expensive way to do a fairly simple scan of networks and systems that provide some level of comfort that networks and systems are secure.
Today it’s pretty common to hear people say “we’ve got the network covered; now we’re really interested in pursuing our application security.” In 2006 I remember Charlie Johnson, head of the consulting practice at Symantec, talking about apps being the only thing that mattered and that he was thinking of committing the Symantec consulting team to secure application development. He may have just been thinking out loud, but securing applications has become the focus of many IT security groups almost to the exclusion of focusing on risk to the organization.
Don’t get me wrong, application security is a huge problem and it will remain a problem for many years. However, there are many other areas of risk (perhaps greater risk) that cannot be ignored. At the technical level, system security for off-the-shelf software is a persistent problem. Organizations still struggle to patch quickly and there are often systems with exceptions to the patching process that weaken an organization’s domain and system security. While patching is still an issue, the biggest vulnerabilities are found within network and system configurations. In most (90-95%) of our pen tests we find weak configurations that lead to the complete compromise of an environment. In addition, in many organizations, database groups are silo’d off and don’t get the security attention that they need. Because of this, we find an excessive level of insecure configurations, embedded passwords, and inappropriate trust relationships that can lead to compromise.
With all of these technical vulnerabilities, it’s amazing that an even wider security hole can be found within the physical operations, business process, and personnel at organizations. This is still usually the easiest way to break into an organization. Often it’s combined with technical exploits, but social engineering provides an almost failsafe way to get information and access within technology environments.
I don’t think we should reduce our focus on application security – there’s a lot to do there and it will take many years to secure this aspect of IT within organizations. However, I think it’s incredibly important not to lose sight of what constitutes risk. If you really want to understand and reduce IT related risk, you’ve got to look comprehensively at risk within all aspects of your IT environment – process, physical, network, systems, database, and applications. Because while you may not be looking at these things, it’s certain that at some point, someone looking for the easiest way in will be looking at exploiting these weaknesses.