The UK government is advancing a program described as a "murder prediction" tool, which aims to utilize personal data of individuals known to authorities to identify those most likely to commit homicide.

Initially referred to as the "homicide prediction project," its title has since been revised to "sharing data to improve risk assessment." The initiative, rooted in the government's goal to enhance public safety, has drawn criticism from various campaign groups for its potential implications on privacy and societal equity.

The existence of the program was revealed by the advocacy group Statewatch, with specifics about its operations disclosed through documents acquired via Freedom of Information requests.

Concerns have been raised regarding the use of data from individuals not convicted of any crime, including personal histories of self-harm and incidents of domestic abuse.

Government officials have countered these claims, asserting that only data pertaining to individuals with at least one criminal conviction is utilized.

The Ministry of Justice articulates that the research initiative is designed to examine the characteristics of offenders that elevate the risk of committing homicide and to explore innovative data science methodologies in risk assessment.

A spokesperson emphasized that the project aims to bolster evidence-based approaches in understanding serious crime risks and thereby enhance public protection.

Commissioned by the Prime Minister's office during Rishi Sunak's tenure, the project integrates crime data from various official sources, including the Probation Service and information from Greater Manchester Police prior to 2015. The data processed includes personal identifiers such as names, dates of birth, gender, and ethnicity, alongside unique identifiers from the police national database.

Statewatch has highlighted a portion of the data-sharing agreement between the Ministry of Justice and Greater Manchester Police that stipulates different categories of personal data to be exchanged.

These categories encompass not only criminal convictions but also non-criminal victim data, such as age at first victimization and initial police contact.

Additionally, sensitive data categorized under "special categories of personal data" includes health indicators anticipated to possess significant predictive value, including mental health conditions, substance addiction, suicide risk, and self-harm, along with disability status.

Sofia Lyall, a researcher from Statewatch, expressed concern over the implications of the Ministry of Justice's initiative.

She asserted that similar algorithmic crime prediction systems have historically exhibited inherent flaws, potentially exacerbating existing biases within the criminal justice system.

Lyall criticized the proposed tool for its potential to reinforce structural discrimination against marginalized communities, particularly racial and economically disadvantaged groups.

A representative from the Ministry of Justice clarified that the project is solely for research and exploratory purposes, built on existing data pertaining to convicted offenders maintained by the HM Prison and Probation Service, along with police records.

The spokesperson indicated that the initiative seeks to enhance the existing risk assessment frameworks already employed by the prison and probation service, assessing whether the incorporation of additional data sources could refine predictive outcomes.

A report detailing findings from the initiative is anticipated to be published in the near future.