Stuxnet. According to the NYT it ranks as “the single most powerful cyberweapon ever deployed.” Stuxnet succeeded where inspections, diplomacy and sanctions so far failed: If only temporarily, it halted the Iranian nuclear program. More recently Stuxnet made the news (again) when it was reported that Russian engineers warned their Iranian customers that turning on the reactor would cause it to explode. Begrudgingly, President Ahmadinejad agreed.

As you read through the stories about Stuxnet (and Google shows nearly 900,000 hits for “Stuxnet” in 0.14 seconds) it is difficult to come up with what this cyberweapon cannot do. The Iranians don’t know how to stop it and despite the typical accusations, they still don’t know who in fact developed it.

Befitting a ninja, Stuxnet arrives to its destination in stealth (apparently, through the banality of a USB drive). Having infiltrated the network, it sets out to find its target, silently navigating through the network, evading numerous detection nets through clever tactics, adaptation and disguise. Stuxnet learns and delivers its disabling payload only upon positive target acquisition. And some of these targets are unusual, thought to be vulnerable only in theory, specifically, Programmable Logic Controllers.

My interest in Stuxnet is driven by its genius-level intelligent architecture. As I think about the autonomous intelligent cyber entity (AiCE) CodeX project and the ambitious computational law missions it is projected to fulfill, Stuxnet’s capabilities are dazzling. Amalgamating the concepts of AiCE, with the architectural principles of UNTAME and Stuxnet, it is possible to envision the genesis of an incredibly sophisticated computational law platform. While the latter two are currently military applications, there is precedent for their permeability into civilian hands (e.g., GPS). We’ll have to wait and see.

But its possible even today to glimpse if even just a tiny fraction of what could be. An AiCE designed in accordance with that amalgamation could easily roam through hundreds of servers, digesting terabytes of legal data. Then take that information and deliver relevant, timely recommendations to human and non-human constituencies (such as other AI Agents). Like Stuxnet and UNTAME, AiCE would also learn from its environment (for more about environmental learning, see here) and adapt to changing requirements concomitantly with different jurisdictions and other mission-influencing dynamic variables. These capabilities would enable sophisticated applications of complexity difficult (today) to imagine. When that happens, it will be very interesting to see what we can come up with.