Author
Asst Prof Jerrold Soh
Organisation/Institution
Singapore Management University, School of Law
Country
SINGAPORE
Panel
Private Law
Title
Generative AI and Private Law Inquiry
Abstract
Lawyers, judges, and scholars are increasingly relying on generative artificial intelligence (genAI) systems, including large language models (LLMs), to assist with legal research, drafting, and analysis. This has prompted significant scholarly interest in how far LLMs technologically can, and normatively should, be used in the legal system. Less attention has been paid to the use of LLMs in legal studies. Specifically, how can and should genAI systems be incorporated into the very process of legal scholarship and academic research? In this paper, we examine how AI could support new methods for inquiry, using private law as a textbook field. Conceptualising LLMs as statistical legal observers, we explore how they can be harnessed to discover new insights about legal inquiry. To this end, our present project pursues three objectives. First, we present a framework for priming LLMs to adopt specified legal perspectives through prompt engineering and retrieval augmentation. As an intuitive and demonstrative starting point, these perspectives are expressed in terms of jurisdiction, namely, the common law of Singapore and of Australia. Secondly, by asking differently primed LLMs to opine on the same legal question, we construct an apparatus for experimental comparative law and seek to demonstrate how it can supplement (not supplant) doctrinal comparative analysis. Specifically, we ask LLMs primed separately with Singapore law and Australian law to attempt a purpose-built hypothetical that exposes tort and contract issues. Thirdly, and more importantly for present purposes, we analyse what the ability of modern LLMs to rapidly construct legal answers from different perspectives could possibly tell us about the nature of private law inquiry.
Biography
Jerrold Soh is an Assistant Professor at Yong Pung How School of Law (YPHSL). He specialises in the relationship between law and artificial intelligence and is centrally interested in how the legal system may not only regulate, but also meaningfully use, artificial intelligence. To this end, he has published research on automated vehicle liability, attribution rules for AI systems, expressing laws as code, automated text classification for court judgments, causal legal text analytics, case law citation networks, and the state of legal innovation in the Asia Pacific region. He regularly speaks at conferences, roundtables, and executive education classes on these topics. At YPHSL, he teaches the law of torts as well as an elective on Law and Technology. He holds degrees in Law and Economics from NUS and an LLM from Harvard Law School.
