This page follows a joint process between three organisations from Europe and the MENA region: EuroMedAI, the German Forum for Ethical Machine Decision Making (DF-EME) and the Young AI Leaders Munich Hub. Each organisation came to frontier AI governance from its own mission and context. In 2026 our work came together through the Swedish Dialogue Institute's Mutual Mentorship Programme, which strengthens EU–MENA cooperation. The process continues at the Swedish Dialogue Institute's Annual Dialogue Forum in Amman in November 2026, and at IASEAI'27 in Paris in February 2027. The page brings together our research and internal discussions. It is work in progress, not a settled position, and we will update it as the process develops.
Most countries in Europe and the MENA region will use frontier AI systems that are built elsewhere, mainly in the United States and China. No single middle power can set the rules for these systems alone. Each one still decides what enters its market, which data centres it hosts and which international agreements it backs, and it can act together with other states. We want to know what these states can realistically require of frontier AI developers, and what they could achieve by coordinating.
Each organisation started from its own work and its own part of the problem.
EuroMedAI is an independent network of researchers, civil society and technologists across Southern Europe, North Africa and the Middle East. We started by measuring the AI divide across the Southern Mediterranean. That led us to ask whether governments are ready and resilient enough to oversee the AI systems they already use. That question brought us to frontier AI.
EuroMedAI's flagship study applies a Tier 0–4 maturity framework to nine Southern Mediterranean countries, using data from the Government AI Readiness Index, ODIN and the OECD AI Capability Indicators. It was presented at the Union for the Mediterranean's 7th Trade and Investment Forum in Barcelona and its 5th Digital Transformation Conference in Antalya.
This EuroMedAI chapter looks at the region's gaps in compute, data governance and skills, and at ways to close them through the Global Digital Compact.
We argued that closing the AI divide is a precondition for global AI governance. We proposed four measures, two of them regional: a shared method for measuring AI maturity, a Minimum Irreducible Capacity floor for every state, a shared compute hub for the Southern Mediterranean and a Mediterranean AI Fund.
EuroMedAI convened a side event at Palexpo during the first session of the Global Dialogue on AI Governance, in partnership with BRISD and the Global Center on AI Governance. It brought together UN officials, researchers, civil society and private-sector participants to discuss public-interest AI. The session is listed in the UN's official programme of side events.
Through the Globethics Emerging Leaders in Ethical AI Governance Fellowship, EuroMedAI worked with five city and regional governments: Barcelona, Montpellier, Tunis, Tetouan and Al Zarqa. We interviewed their policymakers and technical experts. None of the five had a way for civil society or the private sector to oversee AI. They lacked the legal authority, the technical expertise and the capacity. This was also true of the two governments inside the EU, where the AI Act applies. The framework will be presented at the Global Ethics Forum in Bali, 20–22 October 2026.
Governments across the region have adopted responsible AI commitments. The Global Index on Responsible AI 2026 puts binding AI policy coverage at 2.94% in the Middle East and 4.68% in Africa, against 45.77% in Europe. We have proposed a study of citizen-facing public services in Egypt, Jordan and Morocco. It would trace each commitment from the responsible institution to what an affected person can actually do, and identify a minimum set of checks that all three countries could apply.
Through the research group of the Center for AI and Digital Policy, EuroMedAI co-wrote the Tunisia report for the forthcoming 7th edition of the AI Index, Artificial Intelligence and Democratic Values.
This article looks at how Morocco is preparing for the use of AI in its 2026 elections.
The readiness work showed a further gap. Governments that cannot oversee the AI they buy will find it even harder to check the claims of the companies building the most capable systems. To prepare for this question, EuroMedAI runs an internal reading series. It uses the Center for AI Safety's textbook Introduction to AI Safety, Ethics, and Society to build shared foundations in AI safety and to work out how our regional work can help reduce catastrophic risks from advanced AI. The series has held more than ten sessions with 16 members. We are now designing an Arabic-first AI safety course for policy and civil society professionals, with a module on the middle-power question: the governance of compute and data centres, verification, and what Arab and Mediterranean states could realistically push for in multilateral settings.
DF-EME (Deutsches Forum für Ethisches Maschinelles Entscheiden e.V.) is a think tank for the responsible use of algorithms and artificial intelligence, founded in 2020. Its experts from business, politics and science work on the ethical guidelines that should govern machine-aided decisions. Felix Rank, co-founder and board member of DF-EME, regularly gives keynotes and takes part in policy discussions with German and European stakeholders, including the German Federal Ministry for the Environment.
df-eme.orgThis chapter will appear in Governing Artificial Intelligence: Power, Legitimacy, and Democratic Futures (Springer). It looks at who holds power and responsibility when ethical principles for AI are put into practice.
Felix Rank spoke on the Youth stage of the AI for Good Global Summit, convened by the International Telecommunication Union in Geneva. In the panel's session on youth driving inclusive innovation, he joined Young AI Leaders from Africa and Europe to discuss how the benefits of AI can reach everyone, regardless of geography, background or access to technology. The summit took place in the same week as the first session of the UN Global Dialogue on AI Governance, where EuroMedAI held its side event.
Felix Rank is a co-curator of the World Economic Forum's Strategic Intelligence platform for the Future of Work and for Education and Skills. The Future of Work briefing is curated in partnership with him and Prof. Isabell M. Welpe of the Technical University of Munich, who also sits on DF-EME's advisory board. He is a member of the WEF Expert Network and an alumnus of the Global Shapers Community, the Forum's network of young leaders.
The Young AI Leaders Munich Hub grew out of DF-EME in 2023 and is led by Felix Rank as Hub Leader. Since 2025 it has been part of the International Telecommunication Union's AI for Good initiative, in its Young AI Leaders Community. It aims to bring young people's voices into AI policy debates and to support young people around the world in taking a leading role in AI.
munich.young-ai-leaders.orgDuring the World Economic Forum's Annual Meeting, the hub holds a roundtable at AI House Davos, curated by Felix Rank. It brings together young leaders from industry, politics and society to discuss how people around the world can remain active shapers of AI's direction and impact, not only its users. The 2026 roundtable gave rising innovators, including WEF Global Shapers, founders, researchers and industry leaders, a platform to share their experience of AI, from their local realities to ethics and youth-led innovation. They discussed three themes:
An interactive global dashboard that maps where AI's benefits go, such as patents, research publications and corporate headquarters, against its hidden environmental and social costs. It shows at global scale the same divide that EuroMedAI measured in the Mediterranean. It was co-created with the Technical University of Munich and the Munich University of Applied Sciences.
Our three strands came together through the Mutual Mentorship Programme of the Swedish Dialogue Institute for the Middle East and North Africa. The Institute is a Swedish diplomatic mission in Amman whose mandate is to foster dialogue between Europe and the MENA region. The programme pairs about 50 participants from Europe and the MENA region, selected from several hundred applicants, who mentor each other over a year on topics including EU–MENA relations, peace and security, and AI.
Mohamed Alouane (EuroMedAI) and Felix Rank (DF-EME) take part in the 2026 programme. In April 2026 the cohort discussed the risks and opportunities of AI, with a focus on AI for Good. From 17 to 21 May 2026 the cohort met in Runö and Stockholm to discuss the role of AI in EU–MENA cooperation. There they met members of the Swedish Parliament's Committee on Foreign Affairs and Mikael Lindvall, Political Director for Global Affairs at the Ministry for Foreign Affairs. The discussions covered AI for Good, EU–MENA relations in AI development and the role of middle powers. With other EU and MENA participants, Mohamed Alouane also designed a proposal for an AI talent programme linked to the EU AI Act.
The cooperation continued beyond Sweden. In July 2026 both organisations were in Geneva: EuroMedAI convened its side event at the first session of the UN Global Dialogue on AI Governance, and Felix Rank spoke at the AI for Good Global Summit. The next steps are in Amman and Paris.
Sweden is itself a useful example of a middle power in AI. It co-led the UN negotiations on the Global Digital Compact with Zambia, and the Compact created the Independent International Scientific Panel on AI and the Global Dialogue on AI Governance. Sweden does not build frontier models. Its influence on AI comes from leading international processes and from acting through the EU.
Through these exchanges our separate starting points became one shared question: how middle powers in Europe and the MENA region can coordinate on frontier AI governance.
The process continues at the Swedish Dialogue Institute's Annual Dialogue Forum in Amman, Jordan, in November 2026. The forum is the Institute's main event of the year and brings all of its workstreams together. This year it includes an AI for Good track, with about 100 participants from the EU and the MENA region. The forum takes place at a time of conflict and instability in the region, when cooperation between Europe and the MENA region is fragmented.
In the AI for Good track we will hold a joint multistakeholder session with participants from across the EU and MENA region. The session will explore what middle powers can do on frontier AI governance, and how they can coordinate with each other and internationally. Its conclusions will feed into the IASEAI'27 workshop.
IASEAI'27, the conference of the International Association for Safe and Ethical AI at UNESCO House in Paris, is one of several international stages for this work. EuroMedAI is an IASEAI affiliate, and in October 2026 EuroMedAI's chair, Mohamed Alouane, was elected to IASEAI's Working Group for Affiliates. With DF-EME and the Young AI Leaders Munich Hub, EuroMedAI has proposed a half-day workshop for 11–12 February 2027. Before the conference we will mobilise our networks on both shores of the Mediterranean to identify three priority use cases where a middle power needs to set terms for frontier AI it imports. For each use case, a contributor will write a short governance proposal. At the workshop, young AI leaders, Mediterranean civil society and policy researchers, and ethics experts will test each proposal against the questions below.
Within six weeks of the workshop we will publish the stress-test record and the revised proposals on this page. We will then take them to the second session of the UN Global Dialogue on AI Governance in New York in May 2027, and to other forums where middle powers coordinate on frontier AI governance.
You may work on AI policy in Europe or the MENA region, or on frontier AI governance anywhere else. If you would like to suggest a use case, write a proposal or comment on the drafts, contact us or join the network.