AI, Governance Displacement, and the (De)Fragmentation of International Law

31 Pages Posted: 18 Mar 2021 Last revised: 1 Mar 2022

See all articles by Matthijs M. Maas

Matthijs M. Maas

Institute for Law & AI; Centre for the Study of Existential Risk, University of Cambridge; University of Cambridge - Leverhulme Centre for the Future of Intelligence

Date Written: March 17, 2021

Abstract

The emergence, proliferation, and use of new general-purpose technologies can often produce significant political, redistributive, normative and legal effects on the world. Artificial intelligence (AI) has been identified as one such transformative technology. Many of its impacts may require global governance responses. However, what are the direct and indirect effects of AI technologies on the viability, form, or functioning of the international legal order itself? What, if any, are the prospects, peril or promise of AI-driven legal automation at the international level? This paper draws on an ‘AI Governance Disruption’ framework to understanding AI’s impacts on the global governance architecture. Focusing particularly on the potential for legal automation at the international law level, it explores three potential pathways of such ‘legal displacement’: (1) the automation of rule creation and arbitration; (2) the automation of monitoring & enforcement; or (3) the ‘replacement’ of international law with new architectural modes of (international) behaviour control. It then focuses on the effects of these trends on the architecture of international law. It distinguishes 10 different roles that AI applications could play, with distinct effects on the international legal order. That is, AI systems can serve as (1) legal ‘canary in the coal mine’, highlighting the need for greater cross-regime harmonization. However, it can also serve as (2) tough knot or (3) generator of regime fault lines. Under even modest scenarios of legal automation, AI systems may serve variably as a (4) shield, (5) patch, (6) cure, or (7) accelerator of international legal fragmentation. Finally, AI tools may serve as (8) differential enabler; (9) driver of value shifts, or (10) asymmetric weapon, potentially contributing to trends of contestation or erosion in the international legal order. The paper concludes with a brief review of the ways in which international lawyers or regime scholars might approach the risks and opportunities of increasing automation in international law, in order to leverage these trends and tools towards improved efficacy, resilience, and legitimacy of global governance.

Keywords: Artificial Intelligence, AI, International Law, TechLaw, LawTech, Legal Automation

JEL Classification: K33

Suggested Citation

Maas, Matthijs M., AI, Governance Displacement, and the (De)Fragmentation of International Law (March 17, 2021). Available at SSRN: https://ssrn.com/abstract=3806624 or http://dx.doi.org/10.2139/ssrn.3806624

Matthijs M. Maas (Contact Author)

Institute for Law & AI ( email )

Cambridge, MA
United States

Centre for the Study of Existential Risk, University of Cambridge ( email )

Trinity Ln
Cambridge, CB2 1TN
United Kingdom

University of Cambridge - Leverhulme Centre for the Future of Intelligence ( email )

United Kingdom

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
165
Abstract Views
1,065
Rank
306,871
PlumX Metrics