Towards Human-Centered RegTech: Unpacking Professionals' Strategies and Needs for Using LLMs Safely

TL;DR

This paper probes how highly skilled professionals in regulated domains perceive and mitigate compliance risks when using LLMs, revealing data leakage, IP infringement, and output inaccuracy as central concerns. Through 24 expert interviews, it exposes ad-hoc strategies like input distortion and rigorous post-edit verification, underscoring a gap between current NLP tools and frontline compliance needs. The study advances a human-centered RegTech perspective, offering design implications for proactive, governance-aware NLP systems that support professional judgment and accountability. Its contributions provide a bottom-up blueprint for building trustworthy, compliance-driven NLP tools that proactively assist experts in high-stakes workflows, with implications for policy, tooling, and future research. Practically, the work lays groundwork for RegTech technologies that align with real-world practices, risk management, and liability considerations.

Abstract

Large Language Models are profoundly changing work patterns in high-risk professional domains, yet their application also introduces severe and underexplored compliance risks. To investigate this issue, we conducted semi-structured interviews with 24 highly-skilled knowledge workers from industries such as law, healthcare, and finance. The study found that these experts are commonly concerned about sensitive information leakage, intellectual property infringement, and uncertainty regarding the quality of model outputs. In response, they spontaneously adopt various mitigation strategies, such as actively distorting input data and limiting the details in their prompts. However, the effectiveness of these spontaneous efforts is limited due to a lack of specific compliance guidance and training for Large Language Models. Our research reveals a significant gap between current NLP tools and the actual compliance needs of experts. This paper positions these valuable empirical findings as foundational work for building the next generation of Human-Centered, Compliance-Driven Natural Language Processing for Regulatory Technology (RegTech), providing a critical human-centered perspective and design requirements for engineering NLP systems that can proactively support expert compliance workflows.