Cyberwarfare: Play Offense Or Defense?

One of the key differences in military theory between Internet warfare and kinetic warfare is whether defense or offense are stronger. Here's a shortened version of an argument I am formulating about this matter following years of debate.To begin, we require an understanding of theory. While it is a widely held belief that offense is the strongest form of defense, this is unsubstantiated. Military experts worldwide recognize that in the physical world, defense is stronger. Why is this the case, and why doesn't it hold water on the Internet?

There are several reasons, ranging from the defense's ability to choose the battlefield, bring reinforcements, supply the troops, and have knowledge of the land and support of the local population, to the offense's long supply lines, need to defend these, lack of agility on reinforcements, knowledge of the land, and lack of support from the local population.

The defender can also weaken the attacker's position by retreating into his country, making the costs of a successful offense higher.

Then there is attitude and control of momentum, and rarely would you find anyone who disagrees that acting offensively is a bad idea, even in defense.

Carl von Clausewitz said:

"The offensive side can only have the advantage of one complete surprise of the whole mass with the whole, whilst the defensive is in a condition to surprise incessantly, throughout the whole course of the combat, by the force and form which he gives to his partial attacks."

The source of the urban legend on offense being stronger is unclear. In a funsec mailing list discussion some time ago, Tomas L. Byrnes, Gary Warner, and I tried to locate this origin. While most mistakenly attribute it to Clausewitz, many claim it comes from heavyweight fighter Jack Dempsey. As Gary says, "[Dempsey] denies saying this and calls it 'an overworked quotation.'"

Gary pointed out that in Dempsy's 1950 book, "Championship Fighting," he wrote:

"You must have a good defense to be a well-rounded fighter. And the best defense is an aggressive defense."

However, no rule ever holds for all cases, and as Tom pointed out, Clausewitz said:

We must say, that the defensive form of war is in itself stronger than the offensive."

Tom reminds us that Clausewitz wrote this "in an age when the greatest mobility was on horseback," which as Tom mentions, Guderian showed with Blitzkrieg that when you have the advantage of maneuver, you can flank a static defense.

What of Cyber Warfare? Is Defense Still Stronger?
While some lesser-appreciated types of defense work exactly by adding maneuverability to the equation, classic defenses on the Internet don't have much maneuverability at all. The particular for the "real world" becomes the general for the virtual one: Offense is much stronger.

To begin with, we must examine the mindset of both attacker and defender in cyberspace.

Security researchers know for a fact that there are always bugs in programs, and therefore it follows that there are always more vulnerabilities to be found and exploited. The only question is how much time (or resources, in general) one is capable and willing to invest.

Defenders, on the other hand, know for a fact that they can't defend everything. Every computer has thousands of pieces of code running on it, interconnected with various trust levels with different hardware, and even more software. Then there are multiple computers in an organization, and these often rely on one another for essential services, such as authentication. Defenders would examine the entire picture and determine the risk.

Which systems are most at risk? Which are actively threatened? Which matter most? What can be done to reduce the risk for each, and how much would it cost?

The attacker needs only ONE vulnerability to gain access, while the defender needs to actively protect everything, try to limit exposure, and increase the attacker's cost. The two mindsets don't work well together.

And defense, in most cases, is purely reactive: It sets the immobile defenses and then tries to detect breaches of policy. A smart attacker could go completely undetected while the defenders are busy dealing with false positive hits in their warning systems.

While defense decides the layout and landscape, it does not control them in any effective fashion. The attacker needs only to be patient and invest the resources, and a breach is guaranteed. That's why some information is stored separately, and why defenders often speak of layered defenses.

Follow Gadi Evron on Twitter: http://twitter.com/gadievron.

Gadi Evron is an independent security strategist based in Israel. Special to Dark Reading.