As riots and looting continue to flare up in London, a group of online sleuths have gotten together on Google Groups to track down perpetrators . The group makes it clear that is it their intention to use facial recognition to identify the rioters seen in online images. A noble effort on the surface, but it comes with its own set of ethical and practical concerns.
In a lot of ways, this has a distinctly vigilante vibe to it. If this group does get access to a powerful facial recognition tool, there is the possibility of misuse. The group has been discussing the issues involved, but most appear to be going ahead with the plan. One user has offered to build the necessary software suing the Face.API tool along with public images on Flickr and Facebook.
Assuming that this method was 100% accurate, which it likely would not be, there is still the chance of wrongly accusing people. A person might have been caught on film simply trying to get away from the violence, or perhaps just standing and gawking. Surely the Internet would pass the information right to police. A list of possible rioters would never end up posted for all to see. Even if it was, the Internet has a reputation for being totally reasonable... right?