Call for regulation of police prediction algorithms

Incorporating machine-learning into the criminal justice system may have ‘unintended or indirect consequences that are difficult to anticipate’, according to academics calling for measures to regulate AI. 

A study published by the Royal United Services Institute and the Centre for Information Rights, University of Winchester, says that while machine-learning algorithms in policing are in their infancy, there is potential for the technology to do more: ‘The lack of a regulatory and governance framework for its use is concerning.’ 

Several forces are testing or considering tests of algorithm-based artificial intelligence systems to identify crime hot-spots or to predict the likelihood of re-offending. Such initiatives have prompted concerns about the quality of data with which machine learning systems are ‘trained’, and whether their decision-making process will be open to challenge. 

The report argues that ‘it is essential that such experimental innovation is conducted within the bounds of a clear policy framework, and that there are sufficient regulatory and oversight mechanisms in place to ensure fair and legal use of technologies within a live policing environment’.

It recommends that the Home Office sets out clear and appropriate constraints governing how police forces should conduct trials of predictive policing tools.