West Yorkshire Police deploys AI analysis of 999 calls


The force will use a tech system powered by artificial intelligence to record, transcribe and automatically categorise a total of more than 20,000 calls made by the public every week

West Yorkshire Police has implemented an artificial intelligence tool designed to provide analysis and categorisation of incoming emergency calls from the public.

The force receives a total of almost 450,000 emergency calls each year, as well as a further 700,000 non-emergency 101 calls. The newly implemented Post-Call Analysis (PCA) technology will be run from “a police-controlled environment”, where it will be used to record calls, before subsequently providing AI-powered transcripts and “short summaries and automatically categorise calls by topic”.

These categorisations will be made using a model “trained exclusively on police data, and outputs are accompanied by plain language explanations”.

The tool, which will also be used to “extract insight” from previous calls to the force’s control room, is designed to provide insight “around patterns of repeat contact” between the police and individuals. This will be supplemented by “hidden vulnerability detection, identifying indicators of vulnerability that may not be captured in operator notes”, according to a press release form the National Police Chiefs’ Council (NPCC).

PCA will also support compliance by ascertaining whether control-room workers have provided the mandated “crime prevention and forensic preservation advice” to callers.

The tool, which is being implemented widely after the completion of a pilot exercise, was created because, while the force takes an audio recording of all calls, “much of the detail from these calls is not consistently captured in force systems”. This is because the data systems operated by individual police services “typically rely on operator notes rather than the full content of calls”, according to the NPCC, which developed the technology in conjunction with the West Yorkshire Police (WYP).

This lack of reliable data “can limit the ability to analyse demand, identify emerging trends, and connect intelligence across incidents”, the NPCC indicated.

In contrast, during the pilot exercise, the West Yorkshire force force found “improved capture of intelligence”,  and logged a 21% increase in the number of calls where “hidden vulnerability” was detected.


Related content


Assistant chief constable Rob McCoubrey of WYP said: “This capability is still in its early stages, but it is already starting to show how it can help us build a clearer understanding of the calls for service coming into our control room. By securely analysing historic calls in a timely and consistent way, it gives us an additional source of insight that can highlight vulnerability we may otherwise not have spotted and help us understand where further staff support or training might be beneficial. This is about gradually improving the quality of information available to us so we can continue to make informed decisions and provide the best service we can to the public.”

He added: “It is important to be clear that this technology supports our control room professionals – it does not replace their expertise. The skill, empathy and judgement of our staff remain central to every contact we handle. The AI simply assists by presenting information more efficiently.”

The AI system was developed as part of the NPCC’s Digital Public Contact (DPC) programme, which continues to work with other local forces on pilot exercises in which new technology is being trialled.

NPCC contact operations lead assistant chief constable Stuart Hooper said: “This is a genuinely significant moment for public contact policing. Through the DPC programme, we set out to enable forces to adopt technology that works for control rooms, not around them, and seeing Post Call Analysis go live at West Yorkshire Police is a strong example of that approach in action. By capturing and analysing intelligence that has historically been locked away in calls, this capability helps forces better understand demand, identify vulnerability earlier, and make more informed decisions. Just as importantly, it gives time back to control room professionals, allowing them to focus on judgement, empathy and critical decision-making.”

  • This article was amended on 17 March, after the NPCC issued an updated press release clarifying that the PCA tool does not analyse live calls, and that all calls are audio-recorded

Sam Trendall

Learn More →