Sometimes there's a big public "controversy" which is really a non-issue, and "flash mobs" organized via social media belong in this category. The question of the moment is how to combat flash mobs, and there is an obvious answer: mass arrests and prison terms for people who participate in them, combined with efforts to change the social conditions which give rise to violent crime in general. None of this is beyond the ability of democratic societies to understand or manage.
What is causing the flash mobs? It's the confluence of longstanding social problems with new technology that allows virtually instantaneous organization -- and that's pretty much it. Some naïve or disingenuous observers may try to put a "revolutionary" spin on the disorder, but by and large it is simply rank hooliganism. Indeed, there's little that is new in the phenomenon: riots have been the bane of urban societies since the dawn of civilization -- usually involving some incitement by ringleaders or provocateurs, but basically spontaneous in character.
In this context, eliminating the underlying causes of flash mobs will require addressing the persistent, intractable social issues -- poverty, breakdown of respect for authority, glorification of criminality, and so on -- which have always been associated with crime. Given the record of earlier attempts to fight these social maladies, it's reasonable to ask whether this is even possible (it's not like no one ever gave this issue thought before). But as in previous decades, the alternative -- giving up -- is even less attractive, basically guaranteeing that crime and disorder will continue.
From a policing perspective, flash mobs present new but hardly insuperable challenges. As noted, riots have always been more or less spontaneous affairs, making it difficult to predict when or where they will erupt. Previously, the best police could do was react swiftly to stamp out disorder at an early stage, before it spreads -- but social media could actually make it easier to head off disorder before it starts, if police are allowed to monitor social conversations (a concession that looks increasingly likely at this point).
As for people who participate in flash mobs or attempt to incite them, well, there's nothing new here either: you arrest them and put them in jail. This process is facilitated by a number of technological advances, including omnipresent security cameras in businesses and public places; DNA testing; location-tracking of mobile devices; and (fittingly) social media sites which allow law-abiding citizens to help identify miscreants. More thoughtful rioters might try to hide their identities from security cameras by wearing masks or bandanas, but even these won't necessarily be able to defeat increasingly powerful facial recognition software already in use by law enforcement (and the private sector) in the U.S., U.K., and elsewhere.
Here's what won't work: social media blackouts targeting areas or populations affected by civil disorder. In addition to blatantly violating civil liberties, this idea (which is apparently under serious consideration in Britain) is foolish and misguided because it confuses the medium and the message. Does anyone seriously believe that you can prevent angry mobs from forming by blacking out social media? If the government shuts down social media sites, rioters can just resort to more traditional methods of communicating like landlines, graffiti, and direct verbal communication. In fact, it might be smarter to leave social media sites up, in the hope that miscreants will foolishly discuss their plans there, allowing police to head off disorder. In this scenario social media could actually be a tool for police, rather than criminals (at least until the latter get wise to it).