This is the third part of a series of articles that are providing recap of the recently concluded Africa AI Conference held in Kigali in June 2023. In the first part, we broke down the objectives of the conference shared through the powerful keynote address by Nnenna Nwakanma that was entered on the significance of understanding the why behind AI development and the importance of approaching it responsibly. In the second part, we shared the challenges and opportunities that the different stake holders see in the capacity of Africans to harness the power of AI
In this article, we will take a look at AI Governance and Policy
AI Governance and Policy
As artificial intelligence (AI) becomes increasingly pervasive in our lives, the need for comprehensive governance and policies to ensure its fair and ethical development becomes paramount. In recent workshops and sessions focused on AI Governance and Policy, experts and policymakers have gathered to discuss crucial issues concerning equality, inclusivity, and responsible AI deployment. These gatherings have shed light on various challenges faced, potential solutions, and the importance of collaboration in shaping the future of AI.
AI & Equality: A Human Rights Toolbox
One session organized by Women At The Table focused on promoting equality, accountability, and participation in AI and AI products. The workshop emphasized the significance of addressing bias in data and fairness metrics from a human rights perspective. The complex nature of fairness was highlighted, stressing the need to consider context when determining appropriate fairness metrics. Engaging in discussions and workshops on de-biasing data and promoting fairness in AI was encouraged.
Principles of Afro-Feminist AI Data: Visions of Afro-feminist Emancipatory, Liberatory AI
Another session led by Bobina Zulfa from Pollicy, a feminist civic tech organization, explored key issues in promoting inclusivity and addressing biases in AI development. A research paper titled “Afro-feminist AI” discussed the gender data gap, datafication, consent, data governance, and technochauvinism, offering a playbook with draft principles for guiding the design, adoption, and analysis of AI systems. The session emphasized the importance of genuine and accessible participation, recognizing participants, and applying the Maputo Protocol to Afro-feminist AI.
Gender and Inclusion Challenges in AI4D: Learning from Each Other
The session organized by Gender at Work focused on challenges and strategies related to gender inclusion in AI research and organizations. Presenters from AI4AFS and EDUAI Hub shared their insights on gender inclusion in AI for Agriculture and Food Systems and AI models for education, respectively. The breakout room discussions emphasized working at grassroots levels, bridging the gap in AI design, and educating communities about the need for mindset change to promote gender inclusivity.
Related content: EduAI Hub – Advancing Education Innovation Through Responsible AI
Mapping the Networking Landscape for Responsible AI and Gender Equality in Sub-Saharan Africa: A Workshop by the GRAIN Network
The GRAIN network’s session highlighted challenges in the responsible development and deployment of AI innovations in Sub-Saharan Africa. Attendees expressed frustrations with the lack of resources and supportive spaces for women in the AI field. The session emphasized the importance of mapping the AI and gender networks in the region and fostering collaboration to drive innovation and inclusivity.
Crafting User-Centric AI Policies through an Inclusive Process Africa-Asia AI Policymaker Network and Global Partnership for Sustainable Development Data
A panel discussion featuring AI policymakers from Kenya, India, and UNESCO stressed the significance of inclusive AI policy design, addressing ethical concerns, and promoting upskilling and reskilling programs. The experts discussed the role of government in policy-making, collaboration, and self-regulation within the AI community.
Developing Ethical AI Policies in Africa: Co-designing an AI Policy Playbook
Policymakers from Rwanda, Ghana, and Tanzania shared insights into their AI policy development processes. The discussions highlighted the importance of AI policies, including scaling, data availability, and ethical guidelines. The experiences showcased the progress and challenges in developing AI policies in their respective countries.
Jointly Driving Responsible AI Innovation
The Africa-Asia AI Policymaker Network organized a panel session focused on promoting responsible AI innovation and collaboration across different sectors. Participants engaged in discussions and breakout sessions to address key questions related to benefiting vulnerable groups, encouraging collaboration, leveraging generative AI responsibly, and promoting responsible AI innovation without overregulation. The discussions emphasized co-designing solutions with vulnerable groups, ensuring inclusivity, addressing data quality issues, and fostering multi-sectoral collaboration.
Governance in the AI Era: Enabling Digital Transformation
A session led by UNESCO focused on building AI policy in Africa, identifying challenges and the roles of stakeholders in shaping the AI ecosystem. UNESCO’s competence framework highlighted the importance of problem identification, system thinking, and applying digital skills in policymaking. Participants discussed issues such as inadequate infrastructure, lack of AI agenda promotion, limited technical capacity, and data management challenges. Recommendations included decentralized AI initiatives, capacity-building programs, effective policy enactment, and improved data management practices.
Implementation of the UNESCO Recommendation on Ethics in AI in Southern and Eastern Africa
In a panel discussion, policymakers from Rwanda, Namibia, Kenya, and South Africa shared their progress in implementing UNESCO’s AI recommendations. Challenges included limited resources, low digital literacy, and delayed policy enactment. Key opportunities included capacity-building, policy enactment, financing, and collaboration among stakeholders. The session also introduced the Repository Observatory, aiming to facilitate comparative policy learning and showcase distinct AI approaches in different regions.
The MENA AI Observatory: Responsible AI & Data Governance
The session by the American University in Cairo and Birzeit University focused on responsible AI and data governance in the Middle East and North Africa, particularly in agriculture and healthcare. Challenges included limited data accessibility and digitization in healthcare settings. Opportunities identified included converting manual data into digital formats and using AI technologies in healthcare. The MENA AI Observatory was introduced as a platform connecting researchers and innovators in the region, advocating for responsible AI policies.
Researchers: Insights into AI Ethics through an African Lens
Meta Bursaries organized a session for researchers to share findings on the ethical implications of AI application in African institutions. Research projects explored the potential disruption of foreign technology dependency, ethically aligned AI for wildlife conservation, and ethical implications of AI and IoT convergence. Recommendations included increasing awareness of AI systems, reducing foreign dependency, embracing AI in institutions, and integrating AI into educational systems.
The collaborative efforts in these workshops and panel sessions demonstrate a shared commitment to fostering responsible AI innovation in Africa and Asia. Participants have identified challenges, proposed innovative solutions, and emphasized the importance of multi-stakeholder collaboration. By addressing ethical concerns, promoting inclusivity, and developing AI policies aligned with local values and priorities, these efforts will contribute to the responsible and ethical development and deployment of AI technologies, ultimately benefiting societies and advancing digital transformation across the regions.