When video cameras watch and screen: Privacy implications of pattern recognition technologies

Computer vision technologies based on pattern recognition software will soon allow identifying human behaviour that deviates from a pre-defined normality. Such applications are foreseen, amongst others, to be used in public places with purposes of crime prevention, especially in the context of the fight against terrorism. This technology increases the level of automation of video surveillance, changing the main nature of surveillance. The balance of power between the citizen and the State is altered, calling for a new balancing of interests. The automation of risk detection moreover raises the issue of the protection against partially automated decision-making. This paper will deal with the challenges raised by proactive video surveillance technologies to the way how privacy and security have been balanced so far. Attention will moreover be brought to the new safeguards that should be devised to protect the citizens from increased scrutiny and growing automation of the decision-making process.