We recently reported on the U.S. Intelligence Department's plan to peek inside virtual worlds, hoping to develop an automated system that would flag suspicious behavior in the hunt for possible terrorists. At the time, we theorized that it was too lofty a goal, given that just moderating swear words in kids' worlds is an ever-evolving challenge to automate.

Now, an academic has weighed in at Salon; Juan Coles, professor of modern Middle Eastern and South Asian history at the University of Michigan, says that this kind of oversight would be not only infeasible, but unconstitutional:

"Any monitoring by law enforcement of innocuous activity and communication in a virtual world, conducted broadly and without oversight, would be unconstitutional and could invade the privacy of millions of persons. I found no evidence based on my own observations that a virtual world is suitable for planning a terror operation. The anxieties of terrorism "experts" and government officials may also reflect a fear of being not so easily able to extend broad, unchecked surveillance into this new virtual frontier, even if doing so may well accomplish nothing in terms of protecting national security."

The article also suggests that virtual worlds are less than ideal breeding grounds for terrorist groups, given the difficulty in translating virtual weapons and plans into real ones. While it's true that online worlds provide the opportunity for disparate groups to unify and organize, the same could be said for any connected interactive experience -- and one would think that simple email, social networking or simple informational websites would serve as more useful and less transparent communication tools than online worlds. Coles' conclusion? It's government anxiety about a new medium and their poor understanding of it, not an actual terrorist threat, that's at work here.

[Via Eurogamer, Image credit Salon]