Artificial Intelligence (AI) Governance: An Overview

Abstract

The explosion of interest in AI which bestrode the introduction of OpenAI's Chat GPT and GPT 4 underscores the need for this overview to contextualize and address public concerns. If you are like me or most people, my early impression of Artificial Intelligence (AI) was largely influenced and shaped by Hollywood. I remember the doomsday scenario in the Terminator movie franchise. However, there is more to AI than killer robots or drones. AI has huge potential in such areas as finance, agriculture, manufacturing, medicine, robotics, research, education, autonomous vehicles, including law enforcement, military, and defence applications. AI like many human technologies and discoveries has a dual use problem. Meaning it may be employed for good or bad. It is often the fear of the latter that is reflected in movies. The rapid evolution of AI technology at breakneck speed has implications for humans and society. AI and Algorithm bias and discrimination, systemic and environmental risks, the intersections between AI and data privacy, torts, and IP rights violations have made AI governance a necessity and imperative. This paper looks at global and regional efforts to come up with strategies and regulatory frameworks for AI governance. Chief amongst them include the OECD AI Principles; the EU AI Act; and the NIST AI RMF. The common thread running through these frameworks and legislation is identifying and categorizing AI developments and deployments according to their risk levels and providing guidelines for ethical and trustworthy AI with considerations for human safety and innovation. Also identified and examined are a few national and state efforts, namely in the US,UK, Canada, China, Nigeria, and Singapore.

The objective is to facilitate understanding of AI governance and furnish AI developers and deployers with the tools to establish a robust AI risks management framework and compliance regime.

Keywords: AI, AI Governance, Data Privacy, OECD AI Principles, NIST AI RMF, EU AI Act

JEL Classification: K, K13, K33