We were commissioned to evaluate what the optimum policy approach should be for a national registry of police use of Artificial Intelligence (AI). The challenge lay in designing a model that balanced transparency, consistency, and operational practicality across all police forces in England and Wales.

There was no shared understanding of what “AI transparency” should look like in policing — or how to achieve it in a way that built public trust without overburdening forces or compromising sensitive operational information. Our task was to analyse policy options, international precedents, and stakeholder perspectives to define a clear, evidence-based approach to transparency.

We undertook a comprehensive review of AI transparency frameworks, drawing on UK government policy, international registries, and other sectors’ best practice. Alongside this, we engaged with senior policing, policy, and academic stakeholders to understand practical and ethical considerations in publishing information about AI use.

From this, we developed a proposed operating model and governance framework for a future registry — including inclusion criteria, ownership options, and phased implementation recommendations. The work emphasised a narrative-led transparency approach, helping the public understand AI use in plain language, focusing on purpose, outcomes, and safeguards rather than technical detail.

Our findings and recommendations provided a strategic foundation for national policy development on AI transparency in policing. The work informed partners’ thinking on how a public-facing AI registry could operate — enabling greater openness and accountability while maintaining operational integrity.

By combining policy insight, design thinking, and operational understanding, we helped shape a model that supports both responsible innovation and public confidence in the future use of AI across policing.