LLMs & Legal Aid: Understanding Legal Needs Exhibited Through User Queries
This study addresses the problem of understanding user needs in legal aid via LLMs for researchers and practitioners, but it is incremental as it focuses on query analysis rather than model performance.
The paper analyzed user interactions with GPT-4 for legal queries, finding that 1,252 users submitted 3,847 queries, with 70.05% not providing factual information, 64.93% seeking legal information, and 71.43% not imposing requirements on answers.
The paper presents a preliminary analysis of an experiment conducted by Frank Bold, a Czech expert group, to explore user interactions with GPT-4 for addressing legal queries. Between May 3, 2023, and July 25, 2023, 1,252 users submitted 3,847 queries. Unlike studies that primarily focus on the accuracy, factuality, or hallucination tendencies of large language models (LLMs), our analysis focuses on the user query dimension of the interaction. Using GPT-4o for zero-shot classification, we categorized queries on (1) whether users provided factual information about their issue (29.95%) or not (70.05%), (2) whether they sought legal information (64.93%) or advice on the course of action (35.07\%), and (3) whether they imposed requirements to shape or control the model's answer (28.57%) or not (71.43%). We provide both quantitative and qualitative insight into user needs and contribute to a better understanding of user engagement with LLMs.