Durham Constabulary

An early and unusually well-documented custody risk-scoring tool, developed by Durham Constabulary with the University of Cambridge, using random forest models trained on historical criminal justice data to sort individuals into risk categories.

Operational from 2016 until 2021, when Durham Constabulary discontinued it. The stated reason was the resources required to continually refine and refresh the model to satisfy ethical, legal and governance requirements. A Freedom of Information response obtained by Fair Trials showed the tool assessed 12,200 people a total of 22,265 times across that period.

Its use of a commercial postcode-based variable attracted particular criticism. The variable was Experian's Mosaic socio-geodemographic segmentation product, a commercial consumer classification dataset rather than a policing one.

Update, November 2026. The Fair Trials account of its Freedom of Information request (read in full for this review) gives the same figures, 22,265 assessments of about 12,200 people between 2016 and 2021, and quotes Durham Constabulary as saying it stopped using HART "due to the resources required to constantly refine and refresh the model". It reports 12,262 assessments as moderate risk, 7,111 as low and 3,292 as high (which sum to 22,665, not 22,265, so the page is internally inconsistent), and that people rated medium were eligible for the Checkpoint rehabilitation scheme while those rated high were not. The page read carried no publication date. Big Brother Watch (6 April 2018) quotes the force's head of criminal justice saying that one of the variables in the initial model was the Experian Mosaic code, and that the force paid Experian a total of £45,913, including £25,913 for Mosaic. The record's earlier source, the Durham University home page, supports none of this. The original FOI response was not read.