Many security professionals who think they know anything about penetration testing also think they know enough to perform social engineering. After all, they are successful time and time again, so they think they know what they are doing. However, what follows is a textbook example of how a little knowledge in the wrong hands can be very dangerous.I'm referring to the recent case of a social engineering attack against a credit union.
I became famous in the security industry for my social engineering attacks more than a decade ago. Some of the consulting partners I work with get offended when I tell clients that we only use trained intelligence operatives in the performance of any social engineering attacks that I am involved in. I really do only use people trained by intelligence agencies, Special Forces, or similar organizations. I don't approach the task as social engineering, but as an espionage simulation.
My partners get offended because I infer that they are not qualified to do the work, despite the fact that they have successfully performed the work before. The fact is that most professional social engineers are successful not because of their great skills, but because their victims have a gross lack of security awareness. Getting a password from a naive computer user takes no talent whatsoever.
An inexperienced person -- or even any moron -- can be successful 95 percent of the time performing a social engineering attack. The issue, though, is that their results will be much more limited, and there is a distinct possibility that the people can make mistakes and cause damage -- well beyond technical damage.
When these mistakes become public, it hurts the whole industry. This is exactly what happened in the recent case of a security company that used social engineering tactics as part of a penetration test targeting a credit union. The security company created CDs and sent them to employees inside the targeted company. The CDs and associated documents apparently bore the logo of the National Credit Union Association (NCUA), which chastised the credit unionfor allowing the use of the NCUA copyrighted logos and falsifying the signature of its president. I can only hope the security company was appropriately chastised as well.
As this incident became public, some of my potential clients reconsidered social engineering-type tests because of the concerns raised by the credit union case. I can only assume that other security consultants had similar issues. It is very likely that critical work will now not be performed because of this incident.
First, we had a bunch of publicity stunts where security companies publicized how they distributed or sent USB drives to potential victims. In at least one case, the USB drives were distributed without permission, and there was the potential damage to the client's systems. Then people started duplicating this strategy with little thought about possible repercussions.
I have personally heard of police being called, arrests made, security consultants being escorted to the airport, people being fired, etc., all because security consultants think that social engineering is easy and they know what they are doing. I can just imagine the giddiness of the credit union social engineers as they went online and found the NCUA logo and president's signature, and put that on a CD.
I know that many consultants hear me speak and describe my social engineering attacks and think what I do sounds easy enough to replicate or even take a step further. What they don't hear is the years of training I went through, and the preparation to ensure that boundaries are set and enforced. Most important, I am always prepared if I were to get caught.
I will say that there are a significant number of qualified professionals out there with the proper training and experience. But they are in the minority. I know that many of the other people who perform social engineering professionally believe that this article doesn't apply to them, and they know what they are doing. They also think I am completely full of garbage. After all, they haven't been caught and think they would know what to do if they were.
Even if nothing were to go wrong, the quality of the results and the time it takes to get them is a real issue: You can get access to a building, but what difference does that make? You have to be able to prove imminent damage to the client. You can get access to hundreds of computer accounts, but how did you help the client? Proving their employees are naive and their physical security has holes is not a major revelation these days.
Security professionals are here to secure organizations, not confirm they are vulnerable. Any security test has to have an intent of how it should result in improving the organization.
Ira Winkler, CISSP, is president of the Internet Security Advisors Group and author of Spies Among Us. He can be reached via his Website.