Tax Administration - Ministry of Economy and Finance

In my opinion, a successful first Global Dialogue on AI Governance should deliver outcomes that move beyond discussion toward actionable, inclusive, and globally relevant frameworks for cooperation, with particular attention to the needs and realities of the world, including the Global South, and especially Africa. First, it should establish a shared baseline of principles for AI governance that are adaptable to diverse national and cultural contexts. For African countries, this means frameworks that account for varying levels of digital infrastructure, institutional capacity, and data ecosystems, while upholding core values such as fairness, accountability, and inclusiveness. Second, the Dialogue should produce practical and implementable tools, such as policy guidelines, governance toolkits, and capacity-building pathways tailored also to low-and middle-income countries. Supporting African governments in building institutional readiness is critical to ensuring that AI systems are deployed responsibly and effectively. Third, success requires meaningful inclusion of African voices and expertise, ensuring that the continent is not only represented but actively contributes to shaping global norms and standards. Finally, the Dialogue should establish a sustained mechanism for international cooperation, enabling knowledge sharing, technical support, and long-term collaboration.

• AI capacity-building
• Safe, secure and trustworthy AI
• Open-source software, open data and open AI models
• Social, economic, ethical, cultural, linguistic and technical implications of AI

My selection reflects a focus on enabling inclusive, responsible, and sovereign AI development, particularly in the context of African countries. - AI capacity-building is a top priority, as many countries face gaps in technical expertise, infrastructure, and institutional readiness. Strengthening local capacity is essential to ensure that AI systems can be developed, governed, and maintained sustainably, especially that all citizens are starting to use AI tools in their daily daily lives. - Safe, secure and trustworthy AI is critical to building public trust and ensuring that AI systems are reliable, especially in sensitive sectors such as healthcare and public administration, the key takeaway here is that innovation moves faster than trust so we need both to work hand in hand. - The social, economic, ethical, cultural, linguistic, and technical implications of AI are particularly important in diverse contexts such as Africa, where governance approaches must be adapted to local realities. Understanding these dimensions is key to designing inclusive and context-aware AI systems. - Finally, open-source software, open data, and open AI models are crucial for addressing issues of technological sovereignty and reducing dependency on external systems. Open models can enable countries to retain control over their data, foster local innovation, and build AI systems that are aligned with national priorities, while still benefiting from global collaboration.

While the listed themes provide a strong foundation, they remain relatively broad and could benefit from further disaggregation into more specific and actionable areas. Several of these themes encompass complex and multifaceted issues that may require more focused attention. For example, questions related to AI sovereignty and strategic autonomy are not explicitly articulated, yet they represent a major concern for many countries. Issues such as control over data, dependence on external technologies, and the ability to develop and deploy local AI systems are critical dimensions that cut across multiple themes but risk being under-addressed if not clearly identified. Similarly, themes such as "social, economic, ethical, cultural, and technical implications" aggregate a wide range of challenges that may require more targeted discussions to ensure depth and practical outcomes. Further refining these themes into more specific sub-areas would allow for more focused dialogue, clearer priorities, and more actionable recommendations.

In the African context, governance gaps in AI are reflected in several critical challenges. First, AI sovereignty remains a major concern. Many countries rely on external technologies, platforms, and models, which limits control over data, infrastructure, and strategic digital assets. This dependency raises important questions about alignment with national priorities and the ability to develop locally relevant AI systems. Second, there is a clear imbalance between the rapid pace of AI innovation and the slower development of governance mechanisms. AI systems are being adopted across sectors, yet regulatory frameworks, oversight capacities, and institutional readiness are still evolving. This gap creates risks in terms of safety, accountability, and trust. Third, the cultural dimension of AI systems, particularly LLMs, remains insufficiently addressed. Many models are not adapted to local languages, contexts, and societal norms, which limits their usefulness and can lead to misrepresentation or exclusion. Ensuring that AI systems are culturally grounded is essential for their relevance, adoption, and impact.

The AI Dialogue can play a critical role as a neutral, inclusive platform to advance international cooperation on AI governance by bridging diverse perspectives across regions, sectors, and levels of development. It can first facilitate the development of a shared understanding of key principles and priorities, while recognizing the need for flexibility in their implementation across different national contexts. This is particularly important to ensure that countries from the Global South, including Africa, can meaningfully contribute to shaping global norms. Secondly, the Dialogue can serve as a space to translate principles into practical cooperation mechanisms, such as capacity-building initiatives, knowledge-sharing platforms, and jointly developed governance tools. This would help reduce the gap between high-level commitments and real-world implementation. Third, it can promote trust and transparency between stakeholders, including governments, private sector actors, and technical communities. Building trust is essential in a context where innovation is advancing rapidly, and governance mechanisms are still evolving.

In Morocco, for example, the "Maroc AI 2030" strategy reflects a strong national ambition to leverage AI for economic development, innovation, and public sector transformation, while integrating governance considerations. At the international level, Morocco has also been increasingly active in multilateral discussions on AI. As emphasized by Omar Hilale at Gitex Africa 2026, Morocco has the qualifications to serve as a bridge between Africa and global AI governance efforts. This positioning highlights the importance of connecting regional perspectives with global decision-making processes. The added value of the AI Dialogue lies in its ability to link these initiatives into a more coherent and coordinated framework. It can bridge national strategies, such as Maroc AI 2030, with global governance efforts, while ensuring that African countries are not only participants but contributors to shaping international norms.

Governments should play a central role in the AI Dialogue since, usually, the private sector focuses on execution, whereas the government focuses on governance. So I believe the governments are key to this dialogue by sharing national strategies, regulatory experiences, and policy priorities, while also identifying areas where international cooperation is needed. The private sector and technical community can contribute practical insights on implementation, innovation, and emerging risks. Academia and civil society are essential to provide independent research, critical perspectives, and ensure that ethical, societal, and human-centered considerations are fully integrated. To maximize impact, the AI Dialogue should adopt a multi-layered and structured format. First, thematic working groups could be established around priority areas to allow in-depth, technical discussions and the co-development of concrete recommendations. Second, plenary sessions should focus on aligning perspectives, sharing best practices, and building consensus. Third, dedicated spaces should be created to ensure the active participation of underrepresented regions, particularly from the Global South. Additionally, the Dialogue should include mechanisms for continuity, such as follow-up working groups, periodic reporting, and knowledge-sharing platforms. This would ensure that discussions translate into actionable outcomes.

First, African and all other low-and middle-income countries are often underrepresented in shaping global AI norms, despite being increasingly affected by AI deployment. Second, local communities and end-users, especially those in rural or underserved areas, are rarely included in discussions, even though they are directly impacted by AI systems in sectors such as healthcare, public services, and finance. Their lived experiences are critical to understanding real-world implications and building trust. Third, linguistic and cultural diversity remains insufficiently represented. Many AI systems and governance frameworks are developed in limited linguistic and cultural contexts. To address these gaps, the AI Dialogue should adopt inclusive participation mechanisms, such as targeted outreach, regional consultations, and support for participation (e.g., funding, language accessibility). It should also create structured channels to integrate local knowledge and community-level feedback into global discussions.

Multi-stakeholder co-creation labs, regional and thematic roundtables, scenario-based simulations, and policy stress tests could be used to explore how governance frameworks perform in real-world situations.

While several high-level frameworks for AI governance exist, a key gap remains in their practical implementation and operationalization. From my experience working on data governance and AI systems in a public sector context, one of the main challenges is the lack of concrete benchmarks and reference models. When attempting to implement governance structures, it is often unclear which approaches are most effective, whether centralized, decentralized, or hybrid models, particularly at the organizational level. This lack of clarity limits the ability to scale governance practices across departments and institutions. Additionally, there is a shortage of documented, real-world case studies demonstrating how AI governance frameworks are successfully applied in practice. As a result, many institutions are developing approaches in isolation, without the benefit of shared lessons learned or validated best practices. Promising approaches to address these gaps include the development of practical governance toolkits, implementation guidelines, and peer-learning platforms where institutions can share experiences and compare models. Open and collaborative platforms could also support the creation of benchmarking standards and repositories of real-world use cases.