In the summer of 2002 I received a phone call from one of Intel’s senior information security experts, Brian Willis. Brian had just returned from an event in Washington D.C. that he was very excited about. Gartner and the U.S. Naval War College had hosted a three-day seminar-style war game called “Digital Pearl Harbor.” The purpose of the war game was to involve industry for the first time in investigating the possibilities for catastrophic attack of and through the U.S. internet system. They had invited a number of private corporations to participate in this new methodology, and Brian attended as Intel’s representative.
At the time I was working on some risk modeling techniques, so Brian figured I’d be interested in what he had learned. He called and started with, “We have to do this!” He described the event and the possibilities he saw for Intel. The event was very successful and provided much valuable information to the sponsors as defenders, but Brian saw a different aspect. As an “attacker” in the game, he saw how easily and dynamically the attackers in cyberspace were able to build their own systems, business as well as technological, and emphasize their own priorities. The visibility that the game gave into this process came as a bit of a surprise to him and other participants, and Brian recognized how valuable this perspective was to understanding risks facing any defender.
So we decided to stage something similar at Intel, but focusing on the attacker viewpoint rather than the defenders. Although this is somewhat different than a classical war game, we kept the basic process (and the name “war game”) to keep it different from other risk assessment methods. It wasn’t easy to come up with our own game. At the time, there was very little about war gaming that wasn’t based on military objectives, and it was almost all from the defender’s point of view. I even called the U.S. Naval War College; they were very interested and supportive but had little they could share. But through the collective effort of many people, by the summer of 2003 we had put together our own Intel Digital Wargame. The game event itself lasted for two days, and involved nearly every Intel business unit organized in six cells spread across three U.S. cities. It was wildly successful, beyond our expectations, and all the participants said it was exhausting but also both the most instructive and the most fun event they had attended in a long time.
Since then, we have conducted a number of smaller games and continue to have good success with the process. Along the way we have refined it, although we consider it still very much a work in progress. The paper published here is a detailed description of our current process. If war gaming sounds interesting to you, or you are already doing something similar, I hope this will be of use to you. In any case, I would like to hear of your thoughts or experiences or best practices in this area, as we are always looking to learn and improve.