Isabel Tarling, PhD

🚨 [AI GOVERNANCE] The @ai_observatory published the report "Responsible AI Governance in Africa: Prospects for Outcomes Based Regulation," and it's a must-read for everyone interested in global AI governance challenges. Quotes: "At an African level, fewer than 10 countries have adopted a national AI strategy. The common objectives of these AI country strategies include a focus on governance through the establishment of national AI councils, national data strategies, investments in infrastructure, skills and capacity building as well as research and development. Crucially, these strategies offer no insights on what the governance of AI may look like in the future. Further, they offer no clarity around the kind of socio-economic outcomes that the governance must help produce. Consequently, the first pillar in the AU continental strategy is harness human capital development for AI and a strong desire for African governments to collaborate for AI to empower Africa to overcome developmental challenges and drive socio-economic transformation." (page 10) - "While AI policies are emerging across Africa and countries such as Egypt, Mauritius and Rwanda have published their national AI strategies, Africa remains dominated by foreign technology and the foreign AI companies are not necessarily supporting the realization of the national developmental priorities such as those outlined in the AU’s Agenda 2063. Outcomes-based regulations which are formulated based on local socio-economic realities compel regard for these realities notwithstanding the prescriptions in corporate policies of these companies." (page 19) - "The preferred approach to regulation is changing globally. 73 As countries grapple with the challenge of regulating an industry that is evolving and not fully understood, a spectrum of governance models for AI is emerging. For emerging economies, outcomes-based regulation should be preferred in acknowledgment of the harm that the rapid adoption of unregulated AI-based systems can cause. This paper has highlighted both the benefits and challenges of outcomes-based regulation of AI. Exercising oversight over the use of AI must focus on mitigating the effects of the technology rather than micromanaging the technology itself. This entails an agile approach of setting risk-based performance standards based on consistently current information of the developments and trends in AI." (page 31) ➡ Read the report below. ➡ To stay up to date with the latest developments in AI policy & regulation, join 29,300+ people who subscribe to my weekly newsletter (link below).