Basma Balabel

On Sunday, May 4ᵗʰ 2025, the Access to Knowledge for Development Center (A2K4D) at the American University in Cairo’s Onsi Sawiris School of Business and its flagship initiative MENA Observatory on Responsible AI, held a roundtable discussion titled “MultiStakeholder Roundtable on Responsible AI in Egypt.” This was the second edition in a series which commenced on January 14ᵗʰ 2025. The first roundtable, co-hosted by A2K4D, ADSERO - Ragy Soliman & Partners, and APCO, explored Egypt’s proposed AI bill, highlighted key priorities and concerns, and allowed participants to offer recommendations. A key recommendation from the first roundtable was the need to define “responsible AI” and take stock of the existing mechanisms put in place to regulate AI in Egypt, which served as the central purpose of the second edition of the series. 

The second roundtable, held at AUC Tahrir Campus’ Oriental Hall,  brought together a diverse group of experts from the private sector, academia, government, parliament, international organizations, as well as policy makers and lawyers, to explore and develop a collective understanding of responsible AI in Egypt specifically. The discussion involved sharing global, regional, and local definitions and principles of responsible AI, including Egypt’s National AI Strategies (2019 and 2025 editions) and the Egyptian Charter for Responsible AI. Participants also shared their existing responsible AI policies and practices and suggested ways to further advance responsible AI in Egypt.

Defining Responsible AI - the Key Pillars

“Responsible AI” is defined by several different organizations and institutions globally, primarily through sets of principles and pillars. The first session aimed to define what “responsible AI” means in the Egyptian context and examine priority pillars. Participants explored global principles for responsible AI defined by the Organization for Economic Cooperation and Development (OECD), the United Nations Educational, Scientific and Cultural Organization (UNESCO), the United Nations Interregional Crime and Justice Research Institute, the Institute for Ethical AI & Machine Learning, and the African Union’s Continental Artificial Intelligence Strategy. Participants also compared them to Egypt’s own Responsible AI Charter, which emphasizes five core principles: human-centeredness, accountability, fairness, security and safety, and transparency. The charter’s principles were used to guide the conversation and should serve as the base of all future activities by the Observatory pertaining to responsible AI. 

Responsible AI Practices in Egypt

During the first session, Dr. Nagla Rizk, Founding Director of A2K4D and the MENA Observatory on Responsible AI, asked participants how they comply with the five principles of the Egyptian Charter in their work. Dr. Maha Bali, Professor of Practice at AUC’s Center of Learning and Teaching, stated that she makes sure teachers understand how AI works in order to integrate it into their classes and ensure that students have critical AI literacy. From the private sector’s perspective, Ebtehal Bassiouny, Government Affairs Director at Microsoft, highlighted that they follow a risk mitigation approach where they ground themselves in user feedback to see how AI is progressing.

Small and Medium Enterprises (SMEs) also shared their best practices and how they align with the Egyptian Charter. Mohamed Abaza, Head of Growth at Synapse Analytics, highlighted that Synapse has a Machine Learning Operations (MLOps) platform, which flags biases, errors, and ensures compliance, acting as a governance platform. Synapse is also working on offering their services to all segments of society in Egypt by making loans more accessible to underserved communities to ensure fairness. On the other hand, Mahmoud AbdelAziz, CEO at DevisionX, stated that DevisionX’s main pillar is human-centeredness, since they work on making AI development more accessible.

Towards a set of Responsible AI Guidelines

The second session of the roundtable served as a brainstorming exercise to highlight priority pillars to develop a set of guidelines to further implement responsible AI for policy and practice in Egypt, as well as explore ways to assess and monitor progress. 

All participants highlighted the importance of governance and regulation. There was a consensus that a flexible, layered approach - combining soft laws, voluntary codes, and sector-specific regulations - would be more effective than rigid legislation. Dr. Ahmed Daher, Deputy Minister at the Ministry of Education, suggested developing a certification system similar to International Organization for Standardization (ISO) standards for responsible AI implementation. On the other hand, Alaa Mostafa, Senator at the House of Representatives, suggested creating a sandbox for companies to test their models and see if their technology abides by the responsible AI principles.

To ensure AI serves all people, education and awareness were recognized as critical pillars. From integrating AI literacy into school curricula to launching nationwide awareness campaigns, participants emphasized the need to build a culture that understands and values responsible AI. Dr. Maha Bali is currently teaching an AI course at the AUC’s Center for Learning and Teaching that educates students on issues related to understanding and identifying bias, interrogating the validity of AI outputs, as well as informing them on what opportunities AI presents for them in the future job market.

Incentivizing Responsible AI Practices

The roundtable proposed innovative ideas to incentivize responsible practices. Nagham ElHoussamy, Associate Director for Research at A2K4D, suggested launching a ‘Responsible AI Pledge’, modeled after the 'Fairwork Pledge' by the Oxford Internet Institute's Fairwork project, which A2K4D previously collaborated on. This pledge would rate companies on their compliance with responsible AI practices. Based on our experience with the Fairwork project, such ratings can informally incentivize SMEs to improve their practices by encouraging them to boost their scores. 

Another suggestion came from Marwa Soudi, Senior Consultant at the MENA Observatory and Founder of IdeasGym, who recommended establishing a long-term award for SMEs demonstrating ethical AI use. Such initiatives could enhance Egypt’s standing in global indices such as the Global Index on Responsible AI (GIRAI), which evaluates local initiatives and could, in turn, motivate the government to prioritize responsible AI.

Next Steps and Proposed Activities for the Observatory

The MENA Observatory was encouraged to continue to serve as a central hub for knowledge sharing in the region by creating a “Responsible AI Library” and mapping Egypt’s AI ecosystem, including businesses, incubators, governmental initiatives, as well as toolkits and other resources on responsible AI. By fostering collaboration among stakeholders and aligning efforts with national strategies, the Observatory can play a transformative role in advancing responsible AI in Egypt and across the region.

In conclusion, the roundtable underscored that responsible AI is not just a technical or legal issue — it is a societal imperative. As Egypt continues to embrace AI, ensuring its ethical, inclusive, and sustainable use will require collective action, continuous learning, and a shared commitment to the public good.