Tuesday, September 16, 2014

We Need More Than Penetration Testing

Last week I read an article titled  People too trusting when it comes to their cybersecurity, experts say by Roy Wenzl of The Wichita Eagle. The following caught my eye and prompted this post:

[Connor] Brewer is a 19-year-old sophomore at Butler Community College, a self-described loner and tech geek...

Today he’s what technologists call a white-hat hacker, hacking legally for companies that pay to find their own security holes. 

When Bill Young, Butler’s chief information security officer, went looking for a white-hat hacker, he hired Brewer, though Brewer has yet to complete his associate’s degree at Butler...

Butler’s security system comes under attack several times a week, Young said...

Brewer and others like him are hired by companies to deliberately attack a company’s security network. These companies pay bounties if the white hackers find security holes. “Pen testing,” they call it, for “penetration testing.”

Young has repeatedly assigned Brewer to hack into Butler’s computer system. “He finds security problems,” Young said. “And I patch them.”

On the face of it, this sounds like a win-win story. A young white hat hacker does something he enjoys, and his community college benefits from his expertise to defend itself.

My concern with this article is the final sentence:

Young has repeatedly assigned Brewer to hack into Butler’s computer system. “He finds security problems,” Young said. “And I patch them.”

This article does not mention whether Butler's CISO spends any time looking for intruders who have already compromised his organization. Finding security problems and patching them is only one step in the security process.

I still believe that the two best words ever uttered by Bruce Schneier were "monitor first," and I worry that organizations like those in this article are patching holes while intruders maneuver around them within the compromised network.

5 comments:

Anonymous said...

Isn't that were a comprehensive security control framework would come into play? There are frameworks like the SANS 20 Critical Security Controls for Effective Cyber Defense, which call out pen testing, patching and monitoring in their controls.

Roberto Martelloni said...

In my opinion, but I doesn't know if someone is already doing that, where possible after a successful break-in during a penetration test, the tester should install a 'network agent' and the organization should measure the time needed by the SOC to detect a potential intruder maneuver around them 'compromised' network.

A sort of 'post exploitation time to detect'. Am I dreaming ?

CT said...

This is unfortunately quite common with both those procuring and those practising. If you ask something along the lines of "Why do you do penetration testing?" the answer in many cases is effectively whac-a-mole.

I'm inclined to say the individual findings/fixes identified by a penetration test are a side product and that a mature organization really uses the results to identify failings in processes which allows them to prevent and detect or deny such things in the future.

By not doing the aforementioned they are only extracting a small portion of the useful information provided by the service they procured. Despite and maybe because of feeling good at hitting some moles, they end up leaving themselves open to getting nailed at a later date by ignoring the rump of information they have in front of them.

Adam said...

"This article does not mention whether Butler's CISO spends any time looking for intruders who have already compromised his organization."

If he's not already detecting and responding to intruders, then he's missing out on all of the free pentests.

ruhul said...

I'm inclined to say the individual findings/fixes identified by a penetration test are a side product and that a mature organization really uses the results to identify failings in processes which allows them to prevent and detect or deny such things in the future.