Making algorithmic decision-making fair

The growing use of machine learning algorithms in policing and the criminal justice system calls to mind the 2002 film Minority Report, in which a police team arrests criminals before they commit their crime. However, if we set aside the implausible nature of this piece of science fiction it still raises important questions about the most appropriate way to use machine learning algorithms to predict future patterns of crime and re-offending without bias. Fortunately, substantial thought is already being given to this question.

The Government announced today that the Centre for Data Ethics and Innovation (CDEI), an independent advisory body set up by Government to investigate and advise on how to maximise the benefits of data-enabled technologies, will be launching an investigation into the potential for bias in algorithmic decision-making, in partnership with the Race Disparity Unit. While the scope of this investigation is wider than the criminal justice system (taking in recruitment, financial services and local government), its work will be welcomed by the many world-leading British data firms operating in this space, who share the Government’s ambition to maximise the benefits of machine learning algorithms while maintaining public confidence in their deployment.

At the recent Home Office official event, Security and Policing 2019, which ADS was proud to support, ADS held a round-table discussion on policing and machine learning algorithms in partnership with the Royal United Services Institute. The discussion focused on the legal, ethical and regulatory challenges posed by the deployment of these algorithms in an operational policing environment, and how to craft a framework that will enable their continuing use. As a RUSI report on this topic notes, police forces in the UK are facing ‘information overload’, overwhelmed by huge amounts of digital data and unable to effectively garner insights to inform their decision making. In this way, machine learning algorithms can offer a cost-effective way to analyse data, identify patterns and predict risk, which is particularly important in a time when police funding is under strain.

ADS represents many British companies working in this space, who have a vital role to play in supporting our law enforcement and criminal justice services in an age of digital and digitally-enabled crime. It is therefore welcome that the CDEI’s investigation will also explore the positive opportunities for these analytical tools to address existing biases. As the Chair of CDEI, Roger Taylor, said, “if we get this right, the UK can be the global leader in responsible innovation.” Industry will wholeheartedly support efforts to achieve fairer decision-making, but it is also important that the Government does not stifle this innovative aspect of the digital economy.