And then there are cases like that of Gary McKinnon, who from his perch in Britain broke into many sensitive defense information systems a decade ago -- often just by searching for points of entry among default passwords that hadn't yet been changed. McKinnon, an autistic man, was looking for the truth about UFOs (who isn't?), and along the way caused some disruption to both Army and Navy systems.
The U.S. government spent years trying to extradite him from the United Kingdom, but the British Home Secretary ruled against the American request last fall on humanitarian grounds (a psych evaluation held that McKinnon would likely kill himself if he were extradited). Still, the charges against McKinnon remain in place, and Washington threatens to keep up the pursuit. But if the notion of trying to attract master hackers to our cause is ever to take hold, this might be just the right case in which President Obama should consider using his power to pardon.
One presidential act of mercy, such as in the case of McKinnon, won't entirely repair relations or build trust between hackers and the government, but it would be a strong signal of officialdom's growing awareness of the wisdom of embracing and employing the skills of these masters of their virtual domain. Over the years I have had the chance to meet and get to know several of the world's very best hackers. What they have in common -- aside from a kind of startling intelligence -- is a deep attraction to the beauty and complexity of cyberspace. They are not motivated by a desire to disrupt; if anything, they are devoted to free, secure flows of information, believing that virtual liberty will often be the herald of freedom in the "real world." One need only look at the antecedents of the Arab Spring to see how close to the truth this view is.
Beyond recruiting IT industry techs and master hackers -- neither of which might, by themselves, fill all Cybercom's needs -- there is one more interesting possibility for filling the ranks: increasing the use of artificial intelligence (AI). The great advantage of AIs -- for the most part, think very bright software, not Robbie the Robot -- is their speed and accuracy. AIs doing good service in the Navy today, for example, include the Aegis ship defense system and the guidance controls for the Tomahawk land-attack missile. The risk in using them without a human in the loop is that they may have poor judgment -- by human standards. So my suggestion is to buddy up AIs with GIs, doubling the force immediately. Smart soldiers paired with smart software. The AIs' quick reflexes could make Secretary Panetta's call for a cyber pre-emption capability a reality, as blocking an attack often requires action in milliseconds. Taking the offensive is slower, given that attacks are usually mounted by surprise. So in this case the human soldier could exercise some control over his desktop buddy.
In sum, the good news is that there are at least three creative ways to begin to realize the vision for the expansion of Cybercom that was recently shared with the public. The bad news is that none of them is being pursued with nearly enough vigor. At a time when others are waging cyberwars -- see the recent reports of Iranian and Chinese cyber operations -- American capacity is growing at a pretty glacial pace. If the fundamental dynamic of the Cold War was the arms race to build nuclear weapons, the driving force in this "cool war" era is an organizational race to build hacker networks.
And so far we have only just laced up our running shoes.