Is Vulnerability Management Broken?

Some argue that it is time to rethink the vulnerability management hamster wheel

Dark Reading Staff, Dark Reading

December 27, 2012

4 Min Read
Dark Reading logo in a gray background | Dark Reading

If the definition of insanity is doing the same thing over and over and expecting different results, then are today's IT security departments just a little nutty about vulnerability management? Some security experts think so, claiming that the vulnerability management process is flawed at most organizations today.

"It's what we call the security insanity cycle," says Anup Ghosh, founder and CEO of Invincea, "which is that cycle of scan your network, patch your vulnerabilities, detect the intrusions, and wash, rinse, and repeat. We keep doing it thinking that somehow this is going to get better, and it doesn't. Vulnerability management as a strategy isn't working -- that method is fundamentally flawed."

[How are CISOs preparing for 2013? See 7 Risk Management Priorities For 2013.]

When most organizations employ vulnerability scanners, the problem is that the scanning itself doesn't necessarily help manage the problem, says Brian Laing, director of U.S. marketing and products for AhnLab, and a long-time figure in the vulnerability management world during stints at ISS and Red Seal Networks. Today's vulnerability numbers are so huge that they make it difficult to relate the knowledge of these problems with actionable work that mitigates the risk around them, he says.

"I've been trying to get that industry to change to a patch view for years," Laing says. "Don't tell me I have 5 million vulnerabilities. Tell me I have to install 1,000 patches. That I can budget for."

But even with a perfect patching and change management routine in place, organizations with enough intellectual property or financial plumbing in place to entice motivated attackers are proving soft targets for targeted attacks.

"If they're targeting you with a piece of malware, they're going to be taking a very structured approach. They're going to come at you with a vulnerability that's not known," Laing says.

One of the application ecosystems that plays out the insanity cycle Ghosh laments is that of Java, which has suffered a seemingly constant volley of zero-day announcements that enterprises can't keep up with on the patch side or through pre-emptive uninstalls.

"You cannot uninstall Java because of the number of enterprise apps that use it," he says. "Then it's why not just patch Java? Well, that's what everyone's trying to do, but every week there's a new exploit for the latest version of Java. You can't stay on top of it."

The way things are trending should be enough to turn security philosophy on its head, he says. The problem is deciding where to turn.

"A lot of people are saying, 'Hey, stop focusing on vulnerabilities because you'll never get your arms all the way around it. Instead, focus on the threats.' Unfortunately, that's where the dialogue stops," says Ghosh, who believes that the threat-centric approach will require architectural creativity.

His firm relies on the containerized approach of application sandboxing, which relies on the same type of internal barriers and chamber doors within the client that make network segmentation such a good idea on the network side.

"So you click on a link and that link happens to infect that application; they're constrained in a container so they can't move outside of that environment to infect the rest of the host and then the network," he says. "That's a different approach to trying to patch everything -- taking an architectural approach to the problem."

Of course, some security experts warn that sandboxing shouldn't be seen as a panacea for the vulnerability management churn. That's because as hackers start tinkering and finding weaknesses in the way the user can "escape" that virtualized container, those barriers could start to vanish.

"Bottom line: The market success of any sandboxing effort will always revolve around how permissions to escape the sandbox are managed," says David Hess, founder of Data Bakery.

That is why Laing still believes a sane patch management cycle does make sense, in concert with a better examination of the threat behavior itself. Behavioral analysis is hardly a new idea, but the methods have room to further evolve. He believes that it might be helpful for security vendors to start thinking of malware a bit like illegal substances. As he explains, when the government first banned certain drugs, like marijuana, it banned a specific chemical compound. But in the case of acid, which could vary wildly in chemical makeup, the ban was on a class of substances that achieved a certain mental state.

"You don't know what it is you're going to look for, so you have to look for the state that it puts the system in rather than just whether it's a specific set of malware," he says. "Look at what it is doing overall. Did it open up the registry? Well, that by itself isn't so bad. Did it write to the registry? That by itself isn't necessarily bad either. Did it write to the run-once registry key? That's bad. String together not just the one behavior, but look at multiple behaviors together."

Have a comment on this story? Please click "Add Your Comment" below. If you'd like to contact Dark Reading's editors directly, send us a message.

About the Author

Dark Reading Staff

Dark Reading

Dark Reading is a leading cybersecurity media site.

Keep up with the latest cybersecurity threats, newly discovered vulnerabilities, data breach information, and emerging trends. Delivered daily or weekly right to your email inbox.

You May Also Like


More Insights