AI4People - On Good AI Governance: 14 Priority Actions, a S.M.A.R.T. Model of Governance, and a Regulatory Toolbox
A I 4 P E O P L E, 2019
46 Pages Posted: 27 Nov 2019
Date Written: November 6, 2019
Abstract
AI4People’s second year of activities has focused on applying – concretely, in real world scenarios and through appropriate governance - those ethical principles of AI announced by AI4People in 2018. The 2019 White Paper gives shape to – whilst establishing priorities and critical issues – 14 Priority Actions, a Model of S.M.A.R.T. Governance and a Regulatory Toolbox, to which governments and businesses alike can refer to – immediately and efficiently.
To conceive the aforementioned, we examine current initiatives and debates on the governance of AI, and consequently provide:
- A definition of the notion of governance and the principles that are at stake in this context
- 14 Priority Actions that can be undertaken immediately, existing within three new groups of priority: forms of engagement; no-regrets actions; and coordination mechanisms for the governance of AI
- A S.M.A.R.T. Model of Governance, for both governments and businesses, adequate for tackling the normative challenges of AI, while being Scalable, Modular, Adaptable, Reflexive, and Technologically-savvy. We call for specific forms of governance that are neither bottom-up, nor top-down, but that are inbetween, and argue that neither co-regulatory models of AI governance - nor forms of self-regulation, nor its variants, such as ‘monitored self-regulation’ – are adequate
- A Regulatory Toolbox, illustrating how the model of S.M.A.R.T. governance works.
Keywords: Artificial Intelligence, Governance, Regulation
Suggested Citation: Suggested Citation