Interface Design to Support Legal Reading and Writing: Insights from Interviews with Legal Experts

2025-09-29

AI Summary

Legal professionals face persistent inefficiencies in reading, writing, and interpreting complex legal documents; existing tools inadequately support cross-document relational reasoning and lack adaptive, human-in-the-loop AI interaction. To address this, we conducted in-depth interviews and task-based observations to systematically characterize real-world workflows and pain points. Building on these findings, we propose a domain-specific framework for elastic AI interaction interfaces in law, whose core innovation is an integrated error-awareness–diagnosis–recovery mechanism. We developed context-sensitive interactive prototypes and empirically evaluated them with legal practitioners. Our study distills key domain-specific requirements and design principles for integrated browsing and drafting assistance, and reveals fundamental limitations of current legal tech tools in semantic relatedness, controllability, and explainability. The work delivers an evidence-driven design guide and technical roadmap for next-generation legal AI systems.

Technology Category

• Humans and AI: Interaction Techniques and Devices
• Philosophy and Ethics of AI: AI & Law, Justice, Regulation & Governance
• Application Domains: Humanities & Computational Social Science

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions

Abstract

Legal professionals spend significant time reading, writing, and interpreting complex documents, yet research has not fully captured how they approach these tasks or what they expect from skimming and writing-support tools. To examine practices and views on emerging tools, we interviewed 22 legal professionals about workflows, challenges, and technology use. In each session, we leveraged prior HCI-based skimming and writing prototypes that surface emergent cross-document relationships and support AI-resilient interaction (noticing, judging, and recovering from model errors or unexpected behavior); participants completed a contextual fit evaluation to assess whether and how they would use the tools, which document types, and at what stages in their work. Our analysis details limitations and challenges in workflows, domain-specific feedback on AI-resilient interfaces, and expert insights on legal tech design. These findings offer actionable guidance for technology designers developing reading and writing-support for legal professionals, and for legal professionals seeking peer-informed tool integration strategies.

Problem

• Analyzing legal professionals' document workflows and technology challenges
• Evaluating AI-resilient interfaces for legal reading and writing tasks
• Developing domain-specific design guidance for legal technology tools

Innovation

• AI-resilient interaction for legal document analysis
• Cross-document relationship visualization for legal reading
• Contextual fit evaluation framework for legal tech