Innovation For Change MENA Hub

From a civil society perspective, and particularly as a practitioner working across the Middle East and North Africa, the first session of the Global Dialogue will be a success if it moves meaningfully beyond consultation theater toward genuine co-governance. First, success requires structured civil society participation — not as a side event or a comment period, but as a substantive voice in shaping outcomes. The Dialogue must establish formal, transparent channels through which civil society organizations, including those from the Global South and conflict-affected contexts, can contribute to working documents and hold deliberations accountable. Second, it must produce regionally relevant outcomes. Governance frameworks developed primarily by technologically advanced economies frequently fail to reflect the realities of the MENA region — including restricted civic space, the use of AI-enabled surveillance against human rights defenders, limited Arabic-language representation in AI systems, and uneven digital infrastructure. The Dialogue must explicitly engage these asymmetries. Third, the session should deliver a concrete accountability mechanism — even a modest one. Civil society organizations in the MENA region regularly document harms from AI-enabled systems used in government surveillance, content moderation, and border control. Affected communities need pathways to raise these harms within global governance processes, not just national ones. Finally, success means the Dialogue produces a working roadmap with civil society checkpoints — milestones that can be independently monitored between sessions. Without this, the July 2026 session risks being remembered as a well-attended conversation that produced no binding change.

• Safe, secure and trustworthy AI
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights
• Transparency, accountability, and human oversight

These priorities directly reflect our experience working within civic space in the MENA region, and our commitment to co-creation and human-centered approaches. Safe, secure and trustworthy AI. The communities we work with are frequently exposed to AI systems they had no role in designing and have no mechanism to challenge. Safety cannot be defined without their input. Social, economic, ethical, cultural, linguistic and technical implications. This theme captures what we see most on the ground. From Arabic language underrepresentation in AI models, to the impact of automation on already fragile labor markets, these implications are real and immediate for MENA communities. Yet their voices are largely absent from the global conversations shaping these systems. Protection and promotion of human rights. Across the region, we have witnessed how AI tools are used to monitor activists, restrict expression, and shrink civic space. Any governance framework that does not explicitly address this is incomplete. Transparency, accountability, and human oversight. This speaks to the core of how we work. Co-creation only means something if the people affected by a decision can understand it, question it, and change it. That principle should apply to AI governance just as it applies to our own practice.

From our experience in the MENA region, the communities most affected by AI systems are also the most excluded from the policies and processes that govern them. This is not just a participation problem. It weakens the quality and legitimacy of governance outcomes. Co-creation and human-centered approaches have demonstrated, in our own work, that involving citizens early and substantively leads to frameworks that are more grounded, more trusted, and more effective. The Dialogue should explicitly address how civil society organizations and citizens can actively shape AI governance, propose alternatives, and hold commitments to account, not only during formal sessions but throughout the inter-sessional process. This requires dedicated mechanisms, not just open doors. Without this, even the most well-designed thematic clusters risk producing frameworks that look inclusive on paper but remain disconnected from the lived realities of the people they are meant to protect.

The MENA region is significantly behind on all four thematic areas we prioritized. Safe and trustworthy AI frameworks are largely absent or unenforced. Human rights protections have not kept pace with AI deployment, particularly in restricted civic spaces. Transparency mechanisms are weak, and the social, cultural, and linguistic implications of AI remain largely unaddressed in regional policy conversations. Civil society organizations and citizens are already navigating the consequences of these gaps, yet remain the least equipped and least included in addressing them. This is precisely the opportunity the Dialogue can seize. Our experience through Innovation for Change's MENA Hub has shown that when citizens and civil society are given the knowledge, tools, and space to engage, they become effective drivers of accountable governance. The challenge is that this capacity remains underdeveloped and under-resourced across the region. Bridging this gap, by preparing and strengthening civil society to take the lead, is not a side objective. It is the foundation for any AI governance framework that is meant to work in the MENA context.

The AI Dialogue has a unique opportunity to do what no existing mechanism has fully achieved: bring together cross-sectoral expertise under one roof, where governments, civil society, the technical community, academia, and the private sector can genuinely join forces rather than operate in parallel silos. From our perspective, three things would make the Dialogue transformative rather than just additive. First, it should establish a structured cross-sectoral platform where diverse expertise is not just represented but actively pooled. AI governance challenges are too complex and too interconnected for any single sector to address alone. Second, it should function as a real support mechanism for all stakeholders, especially those from the Global South and underrepresented regions like MENA, providing the resources, technical assistance, and institutional backing needed to participate meaningfully, not only formally. Third, it should actively expedite knowledge transfer, co-learning, and cooperation across countries and communities. The gap between those shaping AI governance and those living with its consequences is widening. The Dialogue can close it by embedding knowledge exchange into its core design, not treating it as an afterthought.

Maybe the Digital Democracy Initiative by EU and DANIDA

First, communities of practice organized around the core thematic areas. These would allow practitioners, researchers, civil society organizations, and policymakers to build shared knowledge, develop common positions, and feed grounded expertise directly into the Dialogue's formal process between sessions. Second, policy advocacy task forces that bring together cross-sectoral stakeholders to develop concrete recommendations on specific governance challenges. These should have a clear mandate, transparent outputs, and a direct line to the Dialogue's deliberations, ensuring that advocacy is not just heard but reflected in outcomes. Third, sub-regional spaces and working groups that allow stakeholders to engage at a level that reflects their specific contexts before feeding into the global process. For the MENA region, this is particularly important. Global conversations often flatten regional nuance. Dedicated sub-regional spaces would allow civil society and citizens to develop locally grounded positions and arrive at the global table with a stronger, more representative voice.

Some of the most critical perspectives in AI governance are consistently absent from global discussions. Digital rights activists, journalists, and civil society organizations operating at the local level in the Global South are rarely present in spaces like this Dialogue, yet they are often the first to document AI harms and the last to be consulted about solutions. In the MENA region specifically, these actors face a compounded challenge. They operate in restricted civic spaces, with limited resources and institutional support, and in languages underrepresented in both AI systems and governance processes. When global forums are conducted primarily in English and require significant travel, technical familiarity, or institutional affiliation to access, they are structurally exclusive regardless of intent.

I suggest that the project integrates "Learning labs" as a core methodology. As these are structured, hands-on spaces where participants from different sectors and regions work through real governance challenges together, building shared understanding and practical outputs rather than just exchanging statements. In our experience with Innovation for Change, learning labs are particularly effective at breaking down the expertise hierarchy that often silences local and civil society voices in formal settings. Complementing this, peer exchange sessions between practitioners from different regions would allow participants to learn from each other's governance experiences directly, including what has failed, which is often more valuable than showcasing best practices.

As a civil society practitioner working in the MENA region, I do not have specific examples to point to at this stage. This is itself telling. Effective AI governance approaches that are locally owned, rights-based, and inclusive of civil society remain scarce in our context. What we do know, from our work through Innovation for Change, is what good practice looks like in principle: governance that is built with affected communities, that centers human rights, and that gives civil society a genuine role in both design and accountability. Finding and scaling examples that meet this bar, particularly from the Global South, should be a priority task for the Dialogue itself.