Artificial Intelligence (AI) is perhaps the most disruptive and transformative technology of our time. AI has huge implications for humans and society. AI models and AI systems have been deployed in multiple domains like finance, health, argiculture, autonomous vehicles, employment, law enforcement and cybersecurity.
However, just as AI comes with potential benefits it also has potential risks. There are concerns raised about privacy and copyright violations, algorithm bias and discrimination, energy consumption and environmental impact, deepfakes, misinformation and disinformation. It is, therefore, necessary and imperative that a robust AI governance framework is established by AI stakeholders (policy makers and consumers) and AI actors (organizations and individuals active in the design, development and deployment of AI) to promote a humancentric, “innovative” and “trustworthy” AI. AI governance comprises the policies, procedures, processes and standards adopted by an organization to ensure compliance with extant laws and regulations, adherence to core AI principles (like fairness, transparency, accountability, explainability, safety, data privacy and security) assess and manage AI risks to ensure ethical, safe and responsible AI.
Our paper on AI Governance delves into the origins of AI; the importance of AI governance; and the efforts, across jurisdictions, being made by policy makers, academics, civil society and AI actors to establish AI strategies and frameworks to govern the design, development and delployment of AI models and AI systems. The paper examines the OECD AI principles; NIST AI RMF; UNESCO Recommendations on the Ethics of AI; ISO 42001; and the EU AI Act. Country experience(s) are also x-rayed – US. UK, Canada, China, Nigeria, and Singapore.
Recommendations made for assessing and managing AI risks throughout the AI lifecycle. How to set up an AI Oversight Committee, Align AI practices with business goals and objectives; ensure AI literacy (training and awareness programs); and how to develop and implement an AI governance program while leveraging the organization’s existing expertise and GRC architecture.
The objective of the paper is to deepen understanding about AI, its capabilities and limitations, and the global best practices for fostering ethical, innovative, responsible, safe, and trustworthy AI.
The paper is the basis of the upcoming book, AI Governance: A Necessity and Imperative by Alex Wodi and the AI Legislation Tool Kit Project.
WODI & WODI 