The Access to Knowledge for Development Center At AfricAI Conference 2023

Wriiten By: Farah Elbehairy, Alia Shaddad

‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎‎‎‎‎‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎

Leading the conversation on AI for development in Egypt and the region, the Access to Knowledge for Development (A2K4D) hosted a session at the AfricAI Conference 2023: Advancing responsible & open AI ecosystems in Africa. Researchers Alia Shadadd and Farah El Behiry represented the A2K4D at the event, which took place from the 12th - 14th of June, 2023, in Kigali, Rwanda. 

The conference, organized by the International Development Research Centre (IDRC), Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ), and Niyel tackled the theme “Advancing responsible and open AI ecosystems in Africa,” A series of presentations and discussions shed light on Africa’s progress and potential to leverage responsible AI in the fields of education, agriculture, health, policy, governance, as well as gender equity and inclusion. 

A2K4D’s session titled “The MENA AI Observatory: Research Insights & Key Issues Interactive Workshop’’ outcomes of the research project titled “Governing Responsible AI & Data in the Middle East and North Africa,” conducted in partnership with the Center for Continuing Education at Birzeit University in Palestine, and supported by the IDRC. The project aims to strengthen the development, deployment, and governance of responsible AI and data in the MENA region through knowledge creation, networking, policy and regulatory support, interdisciplinary collaboration, and capacity development.

Farah ElBehairy, Senior Researcher at A2K4D, began by presenting preliminary insights about responsible governance of AI and data in the fields of healthcare and food security, from Egypt, Morocco, and Tunisia. ElBehairy provided audiences with background on legal frameworks and development priorities in all three countries, before offering ground-up evidence from ongoing regional case studies. Through a discussion of existing health data governance practices and challenges, her presentation outlined opportunities for more responsible data and AI governance within the regional health sector. 

The session also covered A2K4D’s upcoming research focusing on the governance and use of AI within the overarching objective of food security in the region. The research aims to showcase regional perspectives on the challenges and opportunities AI presents for food security. We hope that the outcomes will help us map out a set of conclusions and policy recommendations for the responsible governance of AI technologies and the right to food in the region.

As discussed in ElBehairy’s presentation, it is evident that, across the MENA region, more can be done to promote responsible AI and data. This includes developing sector-specific data governance regulations and definitions (e.g. what is health and/or agriculture data), rather than broad laws mainly focused on data protection. Additionally , private health institutions stand ahead of public sector ones in terms of health data digitization and AI deployment. Inefficiencies in public data systems exist due to the lack of digitized patient data, limited internet connectivity, inadequate technology infrastructures, as well as a low level of digital literacy among healthcare workers. There are also discrepancies in data practices in urban vs rural healthcare settings.

Alia Shaddad, Researcher at A2K4D, followed with a presentation introducing the MENA AI Observatory, focusing on its ongoing efforts to strengthen the development, deployment, and governance of responsible data and AI in the MENA region. The Observatory aims to serve as a platform that fosters, supports, and expands the existing regional network, explores and influences policies on responsible and inclusive data and AI in the region, and creates a localized body of knowledge on the use of data and AI in the MENA region. 

The presentation emphasized the importance of contributing to the robust global discourse on data and AI from the region by not only portraying a grounded regional perspective, but also, engaging with existing African and global literature. As the conference progressed, multiple conversations with different researchers showed the commonalities that exist between key issues in the MENA region and across the African continent—demonstrating the importance of bringing these different perspectives together to inform one another, unpack commonalities, learn from differences, and explore potential for collaboration. 

Additionally, a significant part of the AfricaAI Conference was discussing language processing models, the availability of data in different African languages, and the different efforts of multiple organizations in that realm. As such, Shaddad proceeded to shed light on the importance of language-related issues in the field of data and AI in the MENA region. Her presentation also outlined the importance of the MENA AI Observatory in this regard: a tool that can be used to leverage data that is in Arabic, as well as AI content for the people of the region.

She elaborated further on the aims of the MENA AI Observatory, which include conducting research from and on the region and creating and supporting a diverse network of researchers, innovators, practitioners, and policymakers. Additionally, this presentation and many others emphasized the importance of localizing knowledge, highlighting the necessity of having a grounded approach that gives voice to local issues instead of adapting to global issues. Finally, the presentation highlighted the Observatory’s intent to utilize local expertise to impact policy-making in the region, and possibly beyond.

Shaddad concluded her presentation with a brief exploration of the use of the design thinking approach to rethink our research processes. This entailed exploring the different facets of design thinking that allow for a more rigorous, non-linear, creative, collaborative (interdisciplinary), grounded, and iterative research process. All attendees were prompted to think about the ways researchers across the African continent can shift from problem-thinking to solution-based thinking, and to adopt relevant methodologies that are contextualized policy-oriented. She wrapped up with an engaging discussion with attendees on the content of the presentation and their own work across the continent, finally leaving us all with extensive insights, and even more to think about within and beyond our own fields.

Among the key takeaways from the conference was the role of African governments in developing our own data sets to make AI work for our low-resource languages as well as responsibly regulating AI and data for AI innovations in Africa. Another key takeaway was the importance of collaboration among AI researchers, AI practitioners, policymakers, and industry to increase the impact of AI and data across Africa.