The use of GenAI by police call handlers to identify self-harm and suicidal ideation in the context of domestic abuse: An exploratory study

Dr. Sam Lewis, University of Leeds with West Yorkshire Police

Artificial Intelligence (AI) has long been a feature of policing. For example, the use of Machine Learning to support predictive analytics (e.g., identifying crime ‘hot spots’), administrative tasks (e.g., automated call transcription), and complex data analysis is well established. The advent of generative AI (GenAI), however, is said to have brought policing to ‘the brink of a once-in-a-generation reform’ (NPCC 2026).

Plans to embed AI across policing include a £115m national centre for AI in Policing (Policing.AI), due to launch in 2026, that will ‘centralise and streamline the development, evaluation and rollout of AI tools’ (NPCC 2026).

Forces at the forefront of these developments are already exploring the strategic use of AI in priority areas. At a time when domestic abuse victim suicides outnumber domestic abuse-related homicides (NPCC 2025), West Yorkshire Police is exploring the use of AI to support early identification of self-harm (SH) and suicidal ideation (SI) among victims of domestic abuse. This could enable faster referrals to life-saving services and reveal the prevalence of SH and SI among DA victims known to police – an ongoing knowledge gap. This collaborative research project will compare the identification of SH and SI amongst DA victims who come to West Yorkshire Police’s attention over a four-month period, by call handlers and an AI model.