Egyptian State Council - Ministry of Justice

A successful first Global Dialogue on AI Governance should deliver more than a general exchange of views. It should produce a practical foundation for sustained international cooperation. First, it should affirm a shared baseline of AI governance commitments: safety, security, trustworthiness, transparency, accountability, human oversight, and respect for human rights and development needs. At the same time, it should clearly recognize that states do not start from the same level of digital readiness, institutional capacity, infrastructure, data availability, or technical expertise. Second, the Dialogue should endorse a development-oriented approach to implementation. Global AI governance will not be credible if it expects uniform compliance from highly unequal ecosystems. Success therefore requires explicit recognition that common commitments must be matched by differentiated implementation pathways and structured international support. Third, the Dialogue should identify concrete support mechanisms for developing countries, including capacity-building, technical assistance, access to compute and cloud infrastructure, support for local-language datasets, institutional training, and financing arrangements aimed at narrowing readiness and digital divide gaps. Fourth, the Dialogue should promote interoperability rather than one-size-fits-all regulation. International cooperation should help countries align around shared principles and compatible governance tools without erasing legitimate policy diversity. Finally, the first Dialogue should establish a forward-looking process: a roadmap for continued engagement, regular reporting, multi-stakeholder participation, and a practical workstream dedicated to implementation challenges in developing countries. In short, success would mean moving from abstract principles to an inclusive and operational framework: one that combines shared governance norms with realistic implementation support and gives developing countries a meaningful place in shaping the future of global AI governance.

• Safe, secure and trustworthy AI
• AI capacity-building
• Transparency, accountability, and human oversight
• Open-source software, open data and open AI models

I selected these four priorities because they are interdependent and together form the minimum foundation for development-oriented global AI governance. Safe, secure and trustworthy AI is indispensable because no meaningful diffusion of AI can occur without public trust, risk management, and safeguards against harm. However, trustworthiness cannot be achieved through principles alone. AI capacity-building is therefore an urgent priority. Many developing countries face major gaps in infrastructure, compute access, cloud services, data readiness, skills, regulatory implementation, and institutional coordination. If these capacity constraints are ignored, global AI governance may widen existing inequalities rather than reduce them. Transparency, accountability, and human oversight are equally essential, especially where AI affects public services, administrative decisions, education, health, employment, or access to opportunities. These elements are necessary to preserve human agency, enable contestability, and ensure that AI remains subject to governance rather than replacing responsibility. I also selected open-source software, open data and open AI models because, when accompanied by appropriate safeguards, they can lower entry barriers, support local innovation, strengthen research, improve linguistic and cultural inclusion, and reduce dependence on a small number of dominant actors. For many countries in the Global South, openness can be a practical pathway to participation and capability development. At the same time, openness should not be treated as unconditional. It should be accompanied by context-sensitive safeguards relating to privacy, safety, security, intellectual property, and misuse risks. Taken together, these four priorities support a balanced model: shared governance principles, meaningful oversight, wider access to AI capabilities, and structured support for countries with differentiated levels of readiness.

In Egypt, and similarly across many developing-country contexts, the most significant governance gap lies not in the absence of policy ambition, but in the distance between emerging governance frameworks and the institutional, technical, and financial capacity required for effective implementation. Egypt has made important progress through its national AI strategy, governance frameworks, ethics and responsible AI guidance, and growing participation in international AI processes. At the same time, national assessments continue to identify gaps in regulatory clarity, data-sharing arrangements, AI procurement, algorithmic accountability, transparency, and implementation readiness across institutions. The most pressing challenges are therefore capacity-related. These include uneven access to cloud and compute infrastructure, limited availability of high-quality and AI-ready data, funding constraints, talent retention difficulties, and persistent digital inclusion gaps, including between urban and rural communities. Such gaps can slow responsible AI adoption, constrain innovation by start-ups and SMEs, and make it harder for developing countries to translate broad governance principles into operational practice. At the same time, the opportunities are substantial. Egypt's experience shows that many countries in the Global South are not passive recipients of global AI norms; they are actively building governance institutions, investing in skills, improving digital infrastructure, and seeking to align AI with inclusive development and public-service improvement. With structured international support—particularly in financing, technical assistance, compute access, data infrastructure, and capacity-building—governance gaps can become a basis for cooperation rather than exclusion. In this sense, the central opportunity is to ensure that global AI governance enables developing countries to participate as contributors, innovators, and co-shapers of trustworthy AI, rather than remaining primarily implementers of standards developed elsewhere.

The AI Dialogue can play a uniquely valuable role by serving as the United Nations' inclusive political platform for connecting AI governance principles with implementation realities. Its added importance lies in its universality: it can bring together governments, international organizations, technical experts, industry, academia, and civil society in a single forum that is explicitly designed to support open, transparent, and inclusive discussions on AI governance. In that sense, the Dialogue should not duplicate existing initiatives, but rather help align them and make them more mutually reinforcing. From a developing-country perspective, the Dialogue can be especially important in advancing a more balanced model of international cooperation—one that combines common governance commitments with differentiated implementation pathways and structured support. It can help ensure that international AI cooperation is not reduced to norm-setting alone, but also addresses the enabling conditions of responsible adoption: financing, capacity-building, access to compute and cloud infrastructure, data readiness, institutional capability, and meaningful participation in standard-setting and governance discussions. This is particularly relevant in contexts such as Egypt, where national strategy and governance commitments already emphasize inclusive AI, regional cooperation, and active engagement in international AI fora, while also recognizing continuing needs in infrastructure, skills, and implementation readiness. The Dialogue can also add value by identifying practical areas for cooperation: interoperable governance tools, public-sector AI safeguards, capacity partnerships, shared technical resources, and follow-up mechanisms that track progress over time. If it succeeds in linking global principles to practical support and inclusive participation, it can become a central forum for turning international AI governance from a fragmented conversation into a more coordinated and development-oriented cooperative process.

The AI Dialogue should build upon and connect with existing international and regional efforts that already provide normative, technical, and implementation-oriented foundations for AI governance. These include, in particular, the Global Digital Compact and the Independent International Scientific Panel on AI within the UN system; UNESCO's Recommendation on the Ethics of Artificial Intelligence and its Readiness Assessment Methodology; the OECD AI Principles, OECD.AI Policy Observatory, and the OECD-GPAI partnership; as well as standards-related work led by ITU, ISO, and IEC. These mechanisms already contribute important building blocks on ethics, evidence, metrics, technical standards, and policy guidance. It should also connect more systematically with regional initiatives, especially the African Union's Continental Artificial Intelligence Strategy, which reflects a development-focused and Africa-centered approach to AI, and similar Arab regional cooperation efforts. This matters because effective global AI governance requires stronger bridges between global frameworks and regional priorities, particularly on capacity-building, digital inclusion, infrastructure, language diversity, and public-interest use cases. Egypt's own AI strategy and national guidelines reflect this multi-layered approach by emphasizing Arab and African cooperation while drawing on UN, UNESCO, OECD, GPAI, and ITU processes. The added value of the AI Dialogue is therefore not to create a parallel governance architecture, but to provide political coherence, inclusiveness, and implementation focus across these efforts. It can connect fragmented initiatives, elevate developing-country priorities, identify cooperation gaps, and encourage practical support arrangements—especially in financing, technical assistance, compute access, and institutional capacity—so that global AI governance becomes more representative, interoperable, and actionable.

Different stakeholders can contribute most effectively if the AI Dialogue is structured not only as a forum for statements, but as a platform for problem-solving and implementation-oriented cooperation. Governments can contribute by identifying policy priorities, governance gaps, and practical implementation needs. International organizations can contribute comparative evidence, technical assistance models, and lessons from existing frameworks. The private sector can provide expertise on safety, testing, infrastructure, and deployment realities. Academia and research institutions can contribute independent analysis, metrics, and evaluation tools. Civil society, affected communities, and professional groups can help ensure that governance remains grounded in rights, inclusion, public trust, and real-world impacts. This multi-stakeholder logic is consistent with Egypt's own strategy and governance documents, which were developed through multi-stakeholder processes and public consultation. In terms of format, the Dialogue would benefit from a layered structure. A high-level plenary should set strategic direction, but it should be complemented by smaller thematic roundtables, regional consultations, and implementation-focused working sessions. Written submissions should be synthesized into concise issue papers in advance, so that discussions can build on evidence rather than repeat general positions. The Dialogue should also include dedicated sessions for developing-country implementation needs, including capacity-building, financing, compute access, and digital inclusion. Egypt's recent experience, including national multi-stakeholder consultations and broader calls for more consultative and collaborative regulation, suggests that structured participation improves both legitimacy and policy quality. A successful structure would therefore combine inclusiveness, regional balance, technical depth, and a clear follow-up mechanism.

Global discussions on AI governance still tend to underrepresent voices from developing countries, especially from Africa and the Arab region, as well as low-resource language communities, rural populations, women, youth, persons with disabilities, smaller innovators, and practitioners working in public-interest sectors. Egypt's own policy and readiness documents illustrate why this matters: they emphasize the need to reflect Arab and African perspectives in global AI governance, to extend opportunities to underrepresented groups, and to address persistent urban-rural, gender, and linguistic inclusion gaps. They also note that minority languages, regional dialects, and historically underrepresented regions remain insufficiently visible in AI development. These voices should not be included symbolically, but through deliberate design choices. First, the Dialogue should ensure balanced geographic representation, with dedicated space for African, Arab, and other Global South perspectives. Second, participation should be supported through travel funding, remote-access options, interpretation, and multilingual documentation, including Arabic and other widely used languages beyond English. Third, the Dialogue should reserve structured speaking opportunities for civil society, youth, women-led initiatives, SMEs, and public-sector practitioners, rather than relying only on state and large-industry interventions. Fourth, written contributions from underrepresented communities should be actively solicited and reflected in official summaries and outcome documents. Meaningful inclusion also requires thematic inclusion. Questions of local language resources, cultural context, digital access, disability inclusion, and implementation capacity should be treated as core governance issues, not peripheral development concerns. A more representative Dialogue would make global AI governance more legitimate, more realistic, and more effective.

The most effective engagement formats would be those that move beyond sequential formal statements and enable practical, cross-sector interaction. One useful format would be moderated problem-solving labs built around concrete themes such as public-sector AI, AI capacity-building, digital inclusion, data governance, or access to compute. These smaller sessions could bring together governments, researchers, companies, and civil society to identify specific governance problems and practical cooperation options. Egypt's own experience with multi-stakeholder consultations, collaborative regulation, and pilot-based approaches suggests that structured dialogue is most productive when it is focused, participatory, and connected to implementation realities. A second valuable format would be regional listening sessions feeding into the global plenary. This would allow African, Arab, and other regional communities to articulate shared priorities before broader negotiation. A third format could involve scenario-based policy exercises, where participants respond to realistic governance cases such as the use of AI in health, education, justice, or public administration. Such formats can reveal tensions between safety, innovation, accountability, and capacity in a more substantive way than general debate. This is particularly relevant in sensitive public-sector contexts, where cautious testing, transparency, and multi-stakeholder review have been recommended. The Dialogue could also use live collaborative drafting sessions for short outcome notes, supported by rapporteurs and digital participation tools that allow real-time input from remote participants. Finally, an ongoing online platform could extend the Dialogue beyond the annual meeting by collecting submissions, tracking emerging issues, and sharing practical tools, case studies, and implementation lessons. In sum, the most effective formats will be interactive, regionally balanced, problem-driven, and designed to convert discussion into practical cooperation.

First, national AI strategies linked to implementation and monitoring can provide an effective governance foundation. Egypt's second National AI Strategy combines development goals with specific governance, data, infrastructure, ecosystem, and talent objectives, and ties them to monitoring and evaluation rather than leaving governance at the level of broad principles alone. Second, institutional responsible -AI frameworks offer a concrete operational model. Egypt's emerging AI regulatory system-built around a responsible AI charter, a proposed AI law, and a Responsible AI Center-illustrates how countries can combine ethical guidance, institutional oversight, and regulatory development in a structured way. Third, organization-level governance tools are especially valuable. Egypt's National Guidelines for Trustworthy and Responsible AI recommend internal AI governance policies, system classification and approval pathways, risk thresholds, data governance rules, testing and audit procedures, human-oversight checkpoints, incident response processes, transparency measures, explanation and appeal mechanisms, and AI-specific procurement safeguards. These are concrete tools that can be adapted across sectors. Fourth, multi-stakeholder readiness assessments and consultations can help countries move from abstract commitments to actionable roadmaps. Egypt's UNESCO RAM process relied on broad stakeholder consultations and interviews across government, academia, civil society, and the private sector, producing a diagnosis of governance gaps and policy priorities. Fifth, collaborative and experimental regulation can support innovation while preserving oversight. Egyptian practice in collaborative regulation, stakeholder consultation, pilot projects, and sandbox-style approaches shows the value of adaptive, evidence-based regulation in fast-moving technological environments. Finally, effective AI governance also requires enabling infrastructure solutions: open government data, AI-ready datasets, cloud and compute partnerships, and capacity-building for public institutions, start-ups, and SMEs. These are essential if governance is to be both responsible and implementable in developing-country contexts.