Our Research

Ethical considerations and legal issues relating to federated learning

Published Date: 01/02/2024

Author(s): Warren B. CHIK, Florian GAMPER

In this chapter, we will explore the global trends and general principles relating to the ethics of federated learning systems. We will also examine the main legal issues relating to federated learning, in particular, the areas relating to the transparency and fairness of AI systems, privacy and personal data protection, intellectual property, legal ownership and rights to the system and output and governance concerns.

Auditing AI for Whom? A Community-Centric Approach to Rebuilding Public Trust in Singapore

Published Date: 23/01/2024

Zhang, Wenxi and Shanmugam, Sharanya and Allen, Jason G and Wong, Willow and Xu, Olivia, Auditing AI for Whom? A Community-Centric Approach to Rebuilding Public Trust in Singapore (December 27, 2023). SMU Centre for AI & Data Governance Research Paper No. 01/2024, Available at SSRN: https://ssrn.com/abstract=4698703 or http://dx.doi.org/10.2139/ssrn.4698703

Deontics and Time in Contracts: An Executable Semantics for the L4 DSL

Published Date: 19/01/2024

Author(s): Seng Joe WATT, Singapore Management University, Oliver GOODENOUGH, Meng Weng (HUANG Mingrong) WONG, Singapore Management University

Existing approaches to modelling contracts often rely on deontic logic to reason about norms, and only treat time qualitatively. Using L4, a textual domain specific language (DSL) for the law, we offer a more operational interpretation of norms, based on states and transitions, that also accounts for the granular timing of events. In this paper, we present a higher-level rendering of the loan agreement from Flood & Goodenough in L4, and an accompanying operational semantics amenable to execution and static analysis. We also implement this semantics in Maude and show how this lets us visualize the execution of the loan agreement.

WATT, Seng Joe; GOODENOUGH, Oliver; and WONG, Meng Weng (HUANG Mingrong). Deontics and time in contracts: An executable semantics for the L4 DSL. (2023). Legal Knowledge and Information Systems: Proceedings of JURIX 2023. 379, 119-124.

Addressing Governance Challenges of Digitalisation and Sustainability: The Case of Central Bank Digital Currency

Published Date: 31/12/2023

Author(s): Heng Wang

Digitalisation and environmental sustainability are widely discussed topics. However, their nexus remains underexplored and can pose significant challenges for governments and industries alike. The environmental implications of digitalisation are becoming increasingly pertinent with the advent of central bank digital currencies (CBDCs) and their inherent energy consumption and production of e-waste. On the other hand, digitalisation could potentially support sustainability efforts. This begs the question of how systems of governance, such as regulatory frameworks and internal organisational governance, should harmonise digitalisation and sustainability goals. Such harmonisation entails ensuring that digitalisation processes are environmentally responsible while exploring how the application and features of digitalisation can help achieve environmental goals. By drawing insights from the case study of CBDCs, this article will utilise the paradigm of adaptive governance to seek potential solutions to the environmental implications of digitalisation.

Governance of Digitalization and Central Bank Digital Currency: Uncertainties, Balancing, and Learning

Published Date: 18/10/2023

Author(s): Heng Wang

Central bank digital currency (CBDC) is a digital version of fiat currency. The governance of CBDC faces uncertainties, many of which are attributable to complexities in technology and regulation. What are the uncertainties faced by digitalization in general, and in particular, CBDC? How might developing approaches like adaptive governance help actors to handle uncertainties? The article seeks to answer these questions and analyse how governance of CBDC and more broadly digitalization should address these uncertainties.

Wang, Heng, Governance of Digitalization and Central Bank Digital Currency: Uncertainties, Balancing, and Learning (September 17, 2023). Available at SSRN: https://ssrn.com/abstract=4574523 or http://dx.doi.org/10.2139/ssrn.4574523

Federated Learning: Revolution or Evolution in Machine Learning? A Legal Analysis

Published Date: 27/09/2023

Author(s): Warren Chik, Florian Gamper

Federated Learning (FL) is a form of machine learning (ML) where algorithms are trained across different data sets, without the sharing of these data sets. The promise of FL is that it enables the training of algorithms without infringing on privacy, even if an algorithm is trained on personal data. There is no lack of articles on FL in the computer science literature, but there is a dearth of writings on, and understanding of, FL among lawyers and legal academics. This article aims to fill the lacuna by providing a comprehensive legal analysis of FL. This article argues that FL should largely be considered outside the material scope of data protection regulation, like the EU’s GDPR or Singapore’s PDPA. However, it will also be argued that it is a mistake to see the potential of FL purely in its application to personal data. As FL may also provide significant benefits when applied to non-personal data. This article will also analyse how the regulatory frameworks for artificial intelligence apply to FL, arguing that they do not pose insurmountable obstacles to FL. Nevertheless, the objective to create ‘ethical’ AI may pose practical challenges for FL. The ultimate purpose of this study is to aid developers who are considering FL to appreciate the legal and regulatory implications of their product, as well to assist legal practitioners and policy makers to evaluate the impact and effects of FL with a view to formulating the appropriate legal response.

Chik, Warren Bartholomew Kam Wai and Gamper, Florian, Federated Learning: Revolution or Evolution in Machine Learning? A Legal Analysis. Available at SSRN: https://ssrn.com/abstract=4585388 or http://dx.doi.org/10.2139/ssrn.4585388

Legal Hypergraphs

Published Date: 12/07/2023

Author(s): Corinna Coupette, Dirk Hartung, Daniel Martin Katz

Complexity science provides a powerful framework for understanding physical, biological, and social systems, and network analysis is one of its principal tools. Since many complex systems exhibit multilateral interactions that change over time, in recent years, network scientists have become increasingly interested in modeling and measuring dynamic networks featuring higher-order relations. At the same time, while network analysis has been more widely adopted to investigate the structure and evolution of law as a complex system, the utility of dynamic higher-order networks in the legal domain has remained largely unexplored. Setting out to change this, we introduce temporal hypergraphs as a powerful tool for studying legal network data. Temporal hypergraphs generalize static graphs by (i) allowing any number of nodes to participate in an edge, and (ii) permitting nodes or edges to be added, modified, or deleted. We describe models and methods to explore legal hypergraphs that evolve over time and elucidate their benefits through case studies on legal citation and collaboration networks that change over a period of more than 70 years. Our work demonstrates the potential of dynamic higher-order networks for studying complex legal systems, and it facilitates further advances in legal network analysis.

Coupette, Corinna and Hartung, Dirk and Katz, Daniel Martin, Legal Hypergraphs (July 5, 2023). Available at SSRN: https://ssrn.com/abstract=4500877 or http://dx.doi.org/10.2139/ssrn.4500877

Defeasible Semantics for L4

Published Date: 28/02/2023

Author(s): Guido GOVERNATORI, Meng Weng (HUANG Mingrong) WONG, Singapore Management University

The importance of defeasibility for legal reasoning has been investigated for a long time (see among other [10, 3, 11]). This notion mostly concerns the issue that textual provisions of (legal) norms typically provide prima facie conditions for their applicability, but to understand a norm in full, we have to evaluate the norms in the context in which the norm is used and to see if other norms prevent it either to apply or to be effective. In other words, when evaluating norms, we must account for possible (prima facie) conflicts and exceptions. Indeed, in general, norms first provide the basic conditions for their applicability. Then, they give the exceptions and exclusions (and they can go on, with exceptions/exclusions of the exceptions/exclusions and so on).

The first issue to address to model legal reasoning is how to model norms. Here, we follow the approach of [12, 4] and stipulate that a norm is represented by an “IF · · · THEN . . .” rule, where the IF part establishes the conditions of applicability of the norm and the THEN part specifies the legal effect of the norm. Where the legal effect of the norm is either that a proposition is taken to hold legally or that a legal requirement (obligation, prohibition, permission) is in force. Moreover, as we have alluded to, the norms are defeasible; thus, the IF/THEN conditional used to model legal norms does not correspond to the material implication of classical logic, and it has a non-monotonic nature. Several approaches have been proposed to reduce or compile the normative IF/THEN conditional. However, in general, as discussed by [13, 8], they suffer from some limitations; for example, the translation to classical propositional logic requires complete knowledge (for any atomic proposition, we have to determine whether it is true or not), it is not resilient to contradictions, and changes to the norms might require a complete rewriting of the translation.

In this work, we are going to examine how to provide an effective and constructive non-monotonic interpretation of (a restricted version of) L4 based on Answer Set Programming (ASP) meta-program. The meta-program gives the semantics of the underlying L4 constructs as well as a computational framework for them.

GOVERNATORI, Guido and WONG, Meng Weng (HUANG Mingrong). Defeasible semantics for L4. (2023). POPL ProLaLa 2023.

Natural Language Processing in the Legal Domain

Published Date: 25/01/2023

Author(s): Daniel Martin Katz, Dirk Hartung, Lauritz Gerlach, Abhik Jana, Michael James Bommarito

In this paper, we summarize the current state of the field of NLP and Law with a specific focus on recent technical and substantive developments. To support our analysis, we construct and analyze a corpus of more than six hundred NLP and Law related papers published over the past decade. Our analysis highlights several major trends. Namely, we document an increasing number of papers written, tasks undertaken, and languages covered over the course of the past decade. We observe an increase in the sophistication of the methods which researchers deployed in this applied context. Slowly but surely, Legal NLP is beginning to match the methodological sophistication of general NLP. We believe this to be a positive trend for the future of the field, but many questions in both the academic and commercial sphere still remain open.

Katz, Daniel Martin and Hartung, Dirk and Gerlach, Lauritz and Jana, Abhik and Jana, Abhik and Bommarito, Michael James, Natural Language Processing in the Legal Domain (January 24, 2023). Available at SSRN: https://ssrn.com/abstract=4336224 or http://dx.doi.org/10.2139/ssrn.4336224

Depending on AI

Published Date: 09/01/2023

Findlay, Mark James, Depending on AI (January 10, 2023). SMU Centre for AI & Data Governance Research Paper No. 1/2023, Singapore Management University School of Law Research Paper , Available at SSRN: https://ssrn.com/abstract=4321460 or http://dx.doi.org/10.2139/ssrn.4321460

Traffic Rule Formalization for Autonomous Vehicle

Published Date: 05/01/2023

Author(s): Hanif Bhuiyan, Guido Governatori, Andry Rakotonirainy, Meng Weng Wong, Singapore Management University, and Avishkar Mahajan

This study devised and implemented a Defeasible Deontic Logic (DDL)-based formalization approach for translating traffic rules into a machine-computable (M/C) format and thus solving rule issues: rule vagueness (open tex-ture expressions) and exceptions in rules. The resulting M/C format of traffic rules can be utilized for automatic traffic rule reasoning to assist the Autonomous Vehicle (AV) in making legal decisions. The method incorporates the compo-nents and behaviour of regulations based on the rule's obligation, prohibition, and permission activities.

The need for the encoding methodology is motivated by the desire for auto-mated reasoning over Autonomous Vehicle information involving traffic rules. A Queensland (QLD) overtaking traffic rule is used as a use case to illustrate this proposed encoding methodology’s mechanism and usefulness.

BHUIYAN, Hanif; GOVERNATORI, Guido; RAKOTONIRAINY, Andry; WONG, Meng Weng; and MAHAJAN, Avishkar. Traffic rule formalization for autonomous vehicle. (2022). LN2FR 2022: Workshop on Methodologies for Translating Legal Norms into Formal Representations in conjunction with JURIX 2022, the 35th International Conference on Legal Knowledge and Information Systems, Saarbrücken, Germany December 14. 1-14.

Driving-Decision Making of Autonomous Vehicle according to Queensland Overtaking Traffic Rules

Published Date: 05/01/2023

Author(s): Hanif BHUIYAN, Guido GOVERNATORI, Avishkar MAHAJAN, Andry RAKOTONIRAINY, Meng Weng (HUANG Mingrong) WONG, Singapore Management University

Making a driving decision according to traffic rules is a challenging task for improving the safety of Autonomous Vehicles (AVs). Traffic rules often contain open texture expressions and exceptions, which makes it hard for AVs to follow them. This paper introduces a Defeasible Deontic Logic (DDL) based driving decision-making methodology for AVs. We use DDL to formalize traffic rules and facilitate automated reasoning. DDL is used to effectively handle rule exceptions and resolve open texture expressions in rules. Furthermore, we supplement the information provided by the traffic rules by an ontology for AV driving behaviour and environment information. This methodology performs auto-mated reasoning on formalized traffic rules and ontology-based AV driving information to make the driving decision by following the traffic rule. The over-taking traffic rule is our case study to illustrate the usefulness of our methodology. The case study evaluation showed the effectiveness of this proposed driving decision-making methodology.

BHUIYAN, Hanif; GOVERNATORI, Guido; MAHAJAN, Avishkar; RAKOTONIRAINY, Andry; and WONG, Meng Weng (HUANG Mingrong). Driving-decision making of autonomous vehicle according to Queensland overtaking traffic rules. (2022). International Workshop on AI Compliance Mechanism WAICOM 2022.

AI employment decision‑making: integrating the equal opportunity merit principle and explainable AI

Published Date: 05/01/2023

Chan, G.K.Y. AI employment decision-making: integrating the equal opportunity merit principle and explainable AI. AI & Soc (2022). https://doi.org/10.1007/s00146-022-01532-w

Legal Dispositionism and Artificially-Intelligent Attributions

Published Date: 04/01/2023

Author(s): Jerrold Soh

It is conventionally argued that because an artificially-intelligent (AI) system acts autonomously, its makers cannot easily be held liable should the system’s actions harm. Since the system cannot be liable on its own account either, existing laws expose victims to accountability gaps and require reform. Recent legal instruments have nonetheless established obligations against AI developers and providers. Drawing on attribution theory, this article examines how these seemingly opposing positions are shaped by the ways in which AI systems are conceptualised. Specifically, folk dispositionism underpins conventional legal discourse on AI liability, personality, publications, and inventions and leads us towards problematic legal outcomes. Examining the technology and terminology driving contemporary AI systems, the article contends that AI systems are better conceptualised instead as situational characters whose actions remain constrained by their programming. Properly viewing AI systems as such illuminates how existing legal doctrines could be sensibly applied to AI and reinforces emerging calls for placing greater scrutiny on the broader AI ecosystem.

Soh, Jerrold, Legal Dispositionism and Artificially-Intelligent Attributions (November 26, 2022). (2023) 43(4) Legal Studies 583-602, Singapore Management University School of Law Research Paper , Legal Studies, volume 43, issue 4, 2023 [10.1017/lst.2022.52], Available at SSRN: https://ssrn.com/abstract=4313467 or http://dx.doi.org/10.1017/lst.2022.52

Compliance through Model Checking

Published Date: 30/11/2022

Author(s): Avishkar MAHAJAN, STRECKER Martin, Seng Joe WATT, Meng Weng (HUANG Mingrong) WONG, Singapore Management University

In this short note, we describe part of a case study about Singapore’s Personal Data Protection Act, which we first presented informally, then formally as interacting Timed Automata. From these, we derive desiderata on a language and verification framework for reasoning about compliance.

MAHAJAN, Avishkar; STRECKER Martin; WATT, Seng Joe; and WONG, Meng Weng (HUANG Mingrong). Compliance through model checking. (2022). International Workshop on AI Compliance Mechanism WAICOM 2022.

ABLI-SAL Cryptoasset Series – Cryptoassets and Property Law (Singapore edition)

Published Date: 31/10/2022

Conversations at the Intellectual Property and Artificial Intelligence Interface: Understanding the Needs of Singapore’s Innovation Community

Published Date: 30/08/2022

Mark Findlay, Ariffin Kawaja, Ong Li Min, Sharanya Shanmugam

Kampung Ethics

Published Date: 10/08/2022

Mark Findlay and Willow Wong

Cryptoassets in Private Law: Emerging Trends and Open Questions from the First 10 Years

Published Date: 31/07/2022

Jason G Allen, Henry Wells, Marco Mauer

Driving Digital Self Determination

Published Date: 30/06/2022

Mark Findlay

Towards a data-driven financial system: The impact of covid-19

Published Date: 30/06/2022

Nydia Remolina Leon

Blockchain Land Transfers: Technology, Promises, and Perils

Published Date: 28/06/2022

Vincent Ooi, Soh Kian Peng, Jerrold Soh, Blockchain land transfers: Technology, promises, and perils, Computer Law & Security Review, Volume 45, 2022, 105672, ISSN 0267-3649, https://doi.org/10.1016/j.clsr.2022.105672.

Models and Data Trade Regulation and the Road to an Agreement

Published Date: 09/06/2022

Henry Gao