Does undercover police work inevitably wrong its targets? Or are undercover activities justified by a general security benefit? In this article I argue that people can make themselves liable to deception and manipulation. The debate on undercover policing will proceed more fruitfully if the tactic can be conceptualised along those lines, rather than as essentially ‘dirty hands’ activity, in which people are wronged in pursuit of a necessary good, or in instrumentalist terms, according to which the harms of undercover work are straightforwardly overcome by its benefits. This article motivates the ‘liability view’ and describes its attractions, challenges, and implications.