Evaluating LLMs for Police Decision-Making: A Framework Based on Police Action Scenarios

TL;DR

This paper introduces PAS, a comprehensive evaluation framework for assessing Large Language Models in police decision-making, highlighting their current limitations and emphasizing the need for reliable AI tools in law enforcement.

Contribution

The paper presents a novel evaluation framework and a large police-related QA dataset, addressing the lack of tailored assessment tools for LLMs in police operations.

Findings

Commercial LLMs struggle with police-related tasks

LLMs often fail to provide fact-based recommendations

The framework enables reliable evaluation of AI in law enforcement

Abstract

The use of Large Language Models (LLMs) in police operations is growing, yet an evaluation framework tailored to police operations remains absent. While LLM's responses may not always be legally incorrect, their unverified use still can lead to severe issues such as unlawful arrests and improper evidence collection. To address this, we propose PAS (Police Action Scenarios), a systematic framework covering the entire evaluation process. Applying this framework, we constructed a novel QA dataset from over 8,000 official documents and established key metrics validated through statistical analysis with police expert judgements. Experimental results show that commercial LLMs struggle with our new police-related tasks, particularly in providing fact-based recommendations. This study highlights the necessity of an expandable evaluation framework to ensure reliable AI-driven police operations. We release our data and prompt template.