The Transformer Led Policing model: a framework for applying generative artificial intelligence in policing

Abstract This article provides two outputs. First, it introduces the Transformer Led Policing (TLP) model, which is a structured framework for integrating Generative Artificial Intelligence (GenAI) in policing. The purpose of developing the framework is to provide a model for police practitioners and researchers considering implementing or studying GenAI to conduct policing functions. Doing so provides consistency and adherence to the European Union’s AI Act and the United Kingdom’s (UK) Covenant for Using Artificial Intelligence in Policing. The TLP framework outlines a three-tier model: Devise, Discuss, Deploy. Each strand includes four components including development of police use cases, prompts, and application programming interfaces, through to implementation and evaluation. To ensure police services adopting GenAI can retain public confidence, the model incorporates components related to stakeholder management regarding legality, security, ethics, and legitimacy. A brief test case is provided to demonstrate the TLPs applicability. Second, the article provides a risk analysis of use cases for academics and practitioners considering the application of GenAI within policing graded against the European Union’s AI Act risk definitions. The use cases presented provide opportunities to apply the TLP and include responding to the public, criminal investigation, intelligence analysis, safeguarding, workforce management, learning and development and administration. The overall intention of the article is to stir debate and discussion about applying GenAI within policing and provide an implementation framework for those considering doing so.