"Spending 90% of the money on things like hygiene and patching is all wrong," Sager said at the federal government's annual Security Automation Conference. "You have to solve the problem at the enterprise level, at the best practices, at the configuration management level. The future of cyber-defense is really about information management."
Sager's organization manages NSA's red and blue teams of security pros that probe and prod government networks for security holes. It also develops NSA security guides, works collaboratively with other agencies on wider security standards, and helps carry out the military academies'' annual Cyber Defense Exercise. Broadly, the group's mission is to look for vulnerabilities and develop security standards.
In his own organization, Sager is trying to drive this point home by teaching red teams that their purpose isn't just to get better at hacking into networks and systems but to translate the problems they find into the language that IT pros and managers use to secure their networks broadly. "We send them in, turn them loose -- and they're really good at what they do, they'll find a problem -- and they’ll tell you exactly how to stop them from coming in next time, but that's not going to solve the problem," he said. "Whatever solution they offer will be undone next patch Tuesday or next time someone twists a knob somewhere."
Sager said that information sharing is vital in terms of protecting networks everywhere. "If it's happening to you, then chances are that it happened to someone else yesterday and will happen to someone else tomorrow," he said, implying that the old norm of hiding successful attacks because of fear of embarrassment or lost business isn't productive. "There aren’t that many new surprises. It's just not communicated well from place to place."
He also said that it's important to think about cybersecurity from frames of reference that typical cybersecurity professionals don't normally think about. "Think about the trouble ticketing system or the licensing system, for example," he said. "If VMs are crashing at a much higher rate than they were last month or obscure versions of programs are being run, might that indicate that something is going on?"
Sager also noted that his closeness to NSA's cybersecurity work enables him to "see a lot of bad guys," and that his observations have led to a few more ideas about how to stop malicious hackers. "The ability to manipulate and raise uncertainty of the bad guy is a critical piece of defending networks going forward," he said. "Our goal is and should be to raise their cost, raise their uncertainty and force them into decisions they don't want to make in order to have them expose themselves.