If You Don’t Hire Robots to Attack Your Networks, You’re Not Doing Security Right

Just outside the door of the Office of the Under Secretary of Defense for Acquisition, Technology, and Logistics, when Frank Kendall and Ash Carter ran the organization in 2008, Kendall put up a sign that said: "In God we trust. All others must bring data."

When you’re dealing with a sprawling military and a complex world, you need real, verifiable data for decisions, whether they are about striking an adversary, investing in a major force or weapons program, or hiring a new senior leader to command an organization. In every case, you need data to make hard choices. The absence of it makes effective management impossible and increases risk. 

For years, the Department of Defense has been trying to elevate its contractors' cybersecurity to prevent network intrusions. The Government Accountability Officer reported this summer that despite major initiatives across the Department, too many major defense programs have often failed to meet their cybersecurity standards. 

No one likes regulations, but they like hostile nation-states poking around national security-related networks and stealing defense information even less. After years of fruitless attempts to persuade government agencies and private firms to better prepare to ward off cyber threats, the arrival of regulations that compel them to do so is a step forward in America’s defense. The regulations go by the name of DoD’s Cybersecurity Maturity Model Certification. In short, they require DoD contractors to follow strict rules designed to protect unclassified information within the DoD supply chain. Contractors who fail to meet the Department’s standards may find themselves denied DoD business. 

Many private companies are understandably daunted by the task of compliance. Here’s where to start: with data. More specifically, chief information security officers need to a way to evaluate their systems and produce real, granular performance data to show security auditors that the company is operating at the level of effectiveness required. 

That requirement will drive many companies toward automated testing. For example, a software platform might draw upon the MITRE ATT&CK framework, a kind of "periodic table" of known threat actors, tactics, techniques and common knowledge of their behaviors. 

The MITRE Corporation, a federally funded non-profit research and development organization working in the public interest, built and released the original ATT&CK framework in 2015 to help defenders all over the world focus on the threats that matter most to cybersecurity. ATT&CK has since gained momentum in the public and private sectors as a globally vetted, one-of-a-kind, all-source repository of adversary behavior. It gives organizations a stable framework against which they can design their defenses.

In using MITRE ATT&CK, a good breach and attack simulation platform should automatically deploy scenarios to painstakingly probe, assess, and validate a company’s cyberdefense capabilities. Today the breach and attack simulation market is a burgeoning, innovative space with a handful of solutions, a few leaders, and competitive pricing for both on-premises and software-as-a-service deployments. 

Automated security testing sharpens organizational defenses and capabilities in a way that manual testing cannot do at sufficient scale. In support of red, blue, and white teams, a testing platform can discover misconfigurations, reveal operator errors, and identify gaps in your defenses. 

By emulating adversary tactics, like lateral movement or privilege escalation, a good breach and attack simulation platform can test best-in-class cyberdefense technologies designed to prevent malware like Sunburst from spreading across a data center, as in the recent SolarWinds intrusion. Ultimately, an automated testing platform generates performance data to help organizations make the most of their valuable defense resources.

This approach can help not just contractors but the Defense Department as well. For example, an automated platform can serve as a “cyberspace operational force” and deploy adversary emulations against elements of the DoD Cyber Mission Force in kinetic or cyberspace-only exercises to test team performance. 

Without continuous, automated testing, DoD and its contractors will remain vulnerable to cyberattacks, their security programs failing silently due to misconfiguration or team performance. With an automated platform, they can improve security postures by focusing people, processes, and security technologies on the threats that matter most.