A new approach to making sense of police data

Policing has never had more data in its reach. From incident logs to intelligence reports, the volume of information generated every day is vast. But as Catherine Levin discovered in the latest episode of For Every Response, the real challenge is making sense of all of it. To explore that challenge, she spoke to Kate Fisher-Stevens, who is helping shape how policing uses data at a national level as the Director of Data and Analytics at the NPCC. 

“Policing’s biggest challenge definitely isn’t a lack of data. We have reams of it,” Kate explains. “But really the challenge is making sense of that data that we already have.” 

One of the biggest barriers is fragmentation. Data sits across 43 forces, in multiple systems, and in different formats. “That makes it difficult to build a unified national picture,” she says. 

Alongside that sits inconsistent data quality. Records can vary, fields may be missing, and much of the information depends on human input. The result is a system where innovation can happen locally, but is difficult to scale.  

“Without consistent standards, it makes it hard to replicate some of the good ideas that are out there in forces.” 

Risk in using data and in not using it 

Data in policing carries inherent risks, but not always in the way people expect. 

“I think there are two sides of risk to data,” Kate explains. “One is risk of its use, but the other is risk of not using it.” 

With so much information available, failing to use it effectively can leave gaps in understanding and decision-making. The challenge is ensuring that data is shared appropriately and used ethically, while still enabling forces to build the clearest possible picture of what is happening. 

Building capability across the workforce 

Making better use of data is a challenge to do with people, not just a technical issue.  Kate describes a tiered approach. Specialist analysts and data professionals are needed to handle complex analysis, while frontline officers must be able to access and act on insights. Leaders should understand enough to steer how data is used. 

“We don’t need all our leaders to be data experts,” she tells Catherine. “But we do need them to be data literate and be able to understand its limitations.” 

That balance is essential if data is to inform decision-making at every level without becoming a barrier to action. 

Ethics at the core 

One theme comes up multiple times during Catherine’s discussion. Ethics form the baseline of the work Kate’s team does.  

“It’s not an add-on to what we do. It’s foundational to our policing legitimacy and our public trust,” Kate says. 

In practice, that means ensuring data use is lawful, proportionate and explainable. It also means recognising and addressing bias. Because policing data is often taken from human input, gaps or inconsistencies can distort the picture it presents. 

“If we have gaps in that data, it’s going to give you a skewed picture,” she explains. “And what we want to do is avoid any kind of inherent bias that may exist.” 

Transparency is also key. Policing must be able to explain how data is used, both within the criminal justice system and to the public. 

Keeping humans at the centre of innovation

As interest in artificial intelligence grows, questions about data quality have become more pressing. But Kate tells Catherine that waiting for perfect data is not the answer. Instead, the focus should be on understanding the strengths and limitations of existing data, improving it over time, and applying the right tools. 

“AI isn’t a solution for every problem,” she adds. “What matters is how those productivity gains translate into a better outcome for the public.” That includes keeping human oversight at the centre. “Making sure humans are really firmly in the loop, and making sure that it’s explainable to the public.” 

Working across the ecosystem 

Delivering that requires collaboration beyond policing. The National Data Analytics Office works closely with partners across government, industry and academia, as well as initiatives such as police-led AI programmes and national problem-solving frameworks. 

Kate emphasises that when it comes to suppliers looking to work with the police, solutions must take policing’s needs into account. “Everything we do in policing has got to be auditable, proportionate and transparent,” she says. 

Building for the future 

Policing’s data challenge is unlikely to disappear and will probably grow as new technologies emerge and public expectations increase. The shift now underway is about building a shared understanding of data as an asset that can improve decision-making, strengthen public trust and support better outcomes on the frontline. 

---

Listen to the full episode below.