Big data policing: Governing the machines?

The chapter will outline recent cases and controversies (such as the use of facial recognition at Notting Hill Carnival and the rise of predictive policing) that illustrate emerging debates about wider matters of privacy and data protection in terms of Big Data, and the concern about disproportionality and ethics. It is argued that these are important matters in their own right, but also because each is related to legitimacy, and so public support, trust and confidence in policing. The chapter explores these debates in three parts:

Governing AI: challenges of regulating software that is opaque and remains the IP of private software corporations; probably hard for non-specialists to understand; self-learning AI is (by definition) hard to regulate once it is initiated;

Privacy: what level of personal information is included? It might be that ‘personal identifiers’ are not included but nonetheless, other characteristics or behaviour are included that actually reveal private information (e.g. that personal health issues can be identified)

Disproportionality: the negative spiral that means those already subject to greater police attention become ever more closely targeted. Not only does this raise problems of equity (and legal problems perhaps) it is also inefficient in terms of tackling crime since it misdirects attention away from swathes of offenders who will never become subject to police attention.

None of these debates arise solely from the growth of big data policing, they have antecedents. However, it is argued that while there are no simple ‘solutions’ to the problems identified it is important to recognise the nature of these challenges such that they do not get lost in the rush to embrace new technologies and practices of policing.