Breaking The Security Fail Cycle

INTEROP -- Las Vegas -- Security's heavy reliance and emphasis on technology--due to both its heritage and the reality of a shortage of manpower--is part of the reason attackers are getting the upper hand, experts said here this week.

A lack of security humans to connect the dots from the abundance of security alerts and data generated by various security tools in the enterprise can easily lead to a needle-in-the-haystack "fail." Target's dismissal of real alerts amid the piles of false positives it had to cull through has become a cautionary tale of just how challenging it is to parse security data today.

Meanwhile, security products and startups continue to emerge and in some cases bolt on yet another layer or specific fix, which complicates things further for the security team.

"When you go walk the [show] floor and see all these new capabilities, and tools that solve this and that … the question you should always ask is 'What is this going to help let me stop doing?'" Mike Zachman, deputy CISO of Caterpillar said today. "Every security solution is an addition, but never a subtraction. You have to make that a mental model: when you buy a tool, what [will it allow you to] stop doing? If you can't answer that, you'd better feel good" about what the product will do, he said during a panel discussion here chaired by John Pironti of IP Architects LLC.

Ofer Elzam, director of enterprise product security at Sophos, says top of mind for him lately is how to make security simpler. "A lot of breaches we see: it's not a lack of tools, but security is very complex. [The victims] have limited resources to manage those," Elzam said.

Meanwhile, companies are wrestling with basic security hygiene duties such as the piles of software patches that keep landing on their desks. Eric Hanselman, chief analyst of 451 Research, says the basics are actually not as easy as they would seem. With patching, for instance, it's not just apply every patch you get. "There are going to be systems we can't patch," where a patch could break an application, he said.

It's a matter of being creative with the resources you have and automating some security processes, said Elliott Glazer of Max Security and a former CISO. "It only takes one unpatched system in a poorly segmented network" for an attack, he said. "You've got to think about ways of mitigating that risk, with people, process, and technology."

When it comes to patching, no company will achieve 100% patched, Hanselman said. "You can't get there. Patch as much as you can," he said. Network segmentation of systems where the "patch density isn't so high" helps as well as monitoring systems that you can't segment, he said.

Caterpillar's Zachman says his company assesses patches for criticality and uses a timetable for applying them. "We have a standard timetable we hold ourselves to, based on criticality," he said. The operational teams who "own" the affected systems are responsible for applying those fixes on those deadlines, he said.

451's Hanselman concurred that the "owners" of vulnerable systems should maintain them and patch them. The vulnerability risk analysis from new patches should be shared with these teams and provided to them, along with an understanding of the risks of the flaws to attack, he said.

This is part of the evolution of the security team as more of a provider to the business, the panelists said.

Caterpillar's Zachman said his team has become akin to security consultants. "We consult with them, explain the risk, and help them understand how they can mitigate it if they can't comply … But it's up to them to say yes, I accept that risk. We make sure it's documented and understood," he said.

"Any discussion in security is a risk-based discussion," Eric Green, senior vice president of Mobile Active Defense, said here today in another panel discussion. "You have to decide where you can prioritize based on the business."