King’s, and The Dickson Poon School of Law in particular, is rapidly becoming a global hub for AI and digital law. There’s a palpable energy here, a recognition that AI isn’t just a new tool but a transformative force for legal scholarship, practice, and societal understanding. My roles, including that of Law School Liaison for the King’s AI Institute's Steering Group and Associate Director of the Centre for International Governance and Dispute Resolution (CIGAD), place me right at the heart of this interdisciplinary ferment. It’s incredibly exciting to be part of an institution that’s not just reacting to change but actively shaping it, fostering collaborations – like mine with Dr Sanjay Modgil from the Department of Informatics – that are essential for tackling complex global challenges.
“The Praxis-Belief Continuum: Revealing Customary International Law through AI-assisted Analysis” is a project designed to tackle one of the most enduring challenges in international law: how to identify customary international law (CIL). CIL is unwritten law derived from the consistent practice of states undertaken with a sense of legal obligation (known as opinio juris). My project aims to develop an AI-powered framework to systematically map the relationship between what states do (their practice) and what they say or believe to be law (their legal arguments and justifications), and how these elements crystallise into binding norms. We’re starting with a Digital Futures Institute Fellowship to build a proof-of-concept, with the ultimate goal of securing an ERC Starting Grant for a full-scale, five-year investigation.
The core problem this project aims to solve is the sheer scale and complexity of identifying CIL. Evidence of state practice and opinio juris is scattered across countless documents, diplomatic exchanges, national legislation, court judgments, and UN debates - in multiple languages and formats. Traditional manual analysis is incredibly time-consuming and often incomplete. This is evident in critical ongoing discussions, like those concerning customary environmental law obligations before the International Court of Justice.
AI, particularly large language models (LLMs), offers a revolutionary solution. These technologies can process and analyse vast quantities of textual data, identify patterns, extract relevant legal arguments, and even map relationships between different actors and their positions. Our project will leverage LLMs to sift through this fragmented evidence, helping us understand how legal norms emerge, evolve, and diffuse across the international system.
While the International Law Commission has provided a methodological framework, the practical application remains a huge hurdle. Our originality lies in using AI to conduct a comprehensive and systematic analysis at a scale previously unimaginable. We’re not just looking at isolated pieces of evidence; we’re aiming to understand the ‘relationality’ of custom - how states influence each other, how ‘discourse coalitions’ form to champion certain norms, and how arguments and actions interact. We plan to develop an open-access, interactive platform. This will democratise access to CIL research, empowering smaller states, academics, and practitioners who currently lack the resources for such extensive analysis.
This collaboration is essential. International law is not straightforward, and legal reasoning is complex. AI experts like Dr Modgil bring the cutting-edge technical expertise in machine learning, logic, and computational argumentation. My legal expertise ensures the AI tools we develop are sensitive to these legal nuances and are grounded in sound legal theory. Neither discipline can solve this alone. For instance, developing approaches to identify opinio juris – the belief that a practice is legally required – needs both an understanding of what constitutes legally relevant belief and the AI techniques to detect expressions of that belief in text. This synergy will allow us to create genuinely innovative methodologies and tools that truly parse legal reasoning.
Academically, we aim to revolutionise how CIL is studied and applied, providing empirical insights into its formation. For AI, the project will advance legal AI by developing novel approaches to argument extraction and norm identification. It will also contribute to computational argumentation by creating new frameworks for understanding legal reasoning patterns based on conduct and justification.
For wider society, the impact could be significant. By making the identification of CIL more transparent and accessible, we can empower smaller states and researchers, improve predictability in international relations, and support evidence-based decision-making by policymakers and international courts. The open-access platform is key to this, fostering a more inclusive and understandable international legal order.
The immediate next step is the 6-month Digital Futures Institute Fellowship, running from February to August 2025. During this pilot phase, we’ll focus on a specific test case area of CIL, compile a focused dataset, develop preliminary AI algorithms, and build a proof-of-concept platform. We’ll be working closely with the Centre for Data Futures, the King’s Institute for Artificial Intelligence, and the Centre for International Governance and Dispute Resolution (CIGAD). The findings and the prototype from this pilot will be crucial for strengthening a larger grant application, which we aim to submit by October 2025. We’re also planning workshops and a symposium to gather feedback and engage with the wider academic and practitioner community.
AI is poised to be a game-changer. Beyond CIL identification, it can enhance treaty analysis, predict dispute outcomes, assist in drafting legal documents, and improve access to justice globally. My work teaching “Generative AI and the Law” at King’s already explores some of these frontiers with students. However, this transformation also brings challenges – questions of bias in algorithms, accountability for AI-driven decisions, and the need for new regulatory frameworks. My research, and the broader work at King’s, aims to not only harness AI’s potential for international law but also to critically engage with these challenges, ensuring that technological advancements serve the cause of justice.