University of Suffolk, UK

A successful first Global Dialogue on AI Governance would be evidenced by concrete, actionable outcomes across AI capacity-building, the multidimensional implications of AI, and the interoperability of governance approaches. In terms of capacity-building, a key outcome would be a shared commitment to advancing inclusive AI literacy through higher education systems. This requires the development of scaffolded training resources designed as open educational resources (OERs), conceived as digital public goods that are accessible, adaptable, and aligned with international frameworks. Such efforts should move beyond technical training to encompass socio-ethical and human rights-based approaches to AI, enabling meaningful public participation. In this regard, alignment with normative instruments such as the UNESCO OER Recommendation and the UNESCO Dubai Declaration would be essential. Regarding the social, economic, ethical, cultural, linguistic, and technical implications of AI, success would involve a clear recognition of inclusion as a cross-cutting priority. This includes addressing gender disparities in AI development, ensuring accessibility and participation for persons with disabilities, and safeguarding the rights and knowledge systems of Indigenous communities. Embedding these considerations within a human rights framework would ensure that AI governance responds to broader societal needs. On interoperability, the Dialogue should produce pathways for aligning national AI strategies with international frameworks, while supporting context-sensitive implementation within higher education institutions. This includes the development of roadmaps for participatory governance in education, guidance for regulatory frameworks, and mechanisms for coordination between AI policy and the higher education sector, such as national AI–higher education partnerships and institutional AI charters. Finally, success would be reflected in strengthened international cooperation, including initiatives that foster multilingual learning, virtual mobility, and cross-border curriculum development, alongside robust ethical AI deployment and data governance frameworks.

• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Interoperability of governance approaches
• AI capacity-building

A successful first Global Dialogue on AI Governance would be marked by tangible progress across three interrelated domains: AI capacity-building, the multidimensional implications of AI, and the interoperability of governance approaches. First, in terms of AI capacity-building, success would entail a clear commitment from Member States to invest in sustainable, equitable, and context-sensitive education and training systems. Crucially, to promote inclusive citizen participation, higher education systems must adopt established frameworks for advancing AI literacy that go beyond abstract principles. These should be operationalised through open educational resources (OER), understood as digital public goods, which are accessible, adaptable, and locally relevant. In alignment with international normative instruments, such resources can support Member States in fostering not only technical competencies but also critical, socio-ethical understanding of AI. Second, the Dialogue should demonstrate a nuanced engagement with the social, economic, ethical, cultural, linguistic, and technical implications of AI. A successful outcome would be the recognition that AI governance cannot be reduced to technical regulation alone, but must address issues such as inequality, cultural diversity, linguistic inclusion, and the protection of fundamental rights. Embedding human rights as a foundational principle across all discussions would be essential. Third, progress on interoperability of governance approaches would be a key indicator of success. This includes fostering coherence between national, regional, and global frameworks, while respecting contextual differences. The Dialogue should facilitate shared principles, common standards, and mechanisms for collaboration, enabling countries to learn from one another and align efforts without imposing uniformity.

Several cross-cutting and emerging issues merit more explicit attention, particularly in relation to inclusive public participation in AI governance. A key consideration is the need to adopt a holistic, rights-based framework such as ROAM-X framework, which foregrounds human rights, openness, accessibility, and multi-stakeholder participation as guiding principles. Within this perspective, inclusion must be understood as a structural and transversal priority. Gender equity remains a pressing concern, as AI systems continue to be disproportionately designed, developed, and governed by men, leading to risks of embedded bias and exclusion. Addressing this imbalance requires not only greater representation of women in AI fields but also the integration of gender-sensitive approaches throughout the AI lifecycle. Equally important is the meaningful inclusion of persons with disabilities. AI systems hold significant potential to enhance accessibility, yet without deliberate design and governance choices, they may reinforce existing barriers. Ensuring accessibility by design, alongside the active participation of persons with disabilities in decision-making processes, is therefore essential. Furthermore, the perspectives and rights of Indigenous communities must be recognised and respected within AI governance. This includes acknowledging issues of data sovereignty, cultural integrity, and the protection of traditional knowledge. Without such considerations, AI risks perpetuating forms of digital colonialism and epistemic injustice.

Governance gaps in AI are significantly shaping the higher education sector across the Mediterranean, revealing both structural challenges and transformative opportunities. As highlighted in the Responsible Artificial Intelligence for Digital Cooperation in the Mediterranean Higher Education report (Atenas & Stefanelli, 2026), a central issue lies in the misalignment between international frameworks and national AI strategies, and their limited translation into actionable pathways within HE systems. While countries are increasingly aligning with guidance from organisations such as UNESCO and the United Nations, implementation within institutions remains uneven, often lacking coordination mechanisms, clear governance structures, and sustained investment. A key challenge is the fragmentation of digital and open knowledge policies. Treating Open Educational Resources (OER), Open Access, Open Data, and Free and Open-Source Software as separate domains constrains the development of integrated digital public goods ecosystems. This fragmentation limits interoperability, reduces efficiency, and undermines the potential of AI to support equitable and scalable knowledge systems. Additionally, many HE institutions face gaps in governance maturity, with limited capacity to move from experimental AI adoption to institution-wide, accountable governance frameworks. At the same time, important opportunities are emerging. AI is acting as a catalyst for institutional transformation, as HEIs can scale innovation in curricula, research, and digital infrastructure. There is also growing momentum around AI literacy, pedagogical innovation, and interdisciplinary collaboration, supported by capacity-building initiatives for both educators and students. Importantly, HE is increasingly positioned as a strategic interface between global digital cooperation and national innovation systems, in this sense, HE requires Digital Education Charters to harmonise ethical principles, data governance standards, and open practices across borders. Strengthening governance through digital public infrastructures and digital public goods will be essential to ensure that AI integration remains ethical, inclusive, and aligned with human rights and sustainable development.

The Global Dialogue on AI Governance can play a key role in advancing international cooperation by acting as a bridge between global normative frameworks, national strategies, and institutional implementation—particularly within the higher education sector. First, the Dialogue can foster alignment between international standards, such as those advanced by UNESCO open education and open science strategies, the United Nations roadmap for digital cooperation, and national AI strategies. By facilitating policy coherence while respecting regional diversity, it can support countries in translating high-level principles into actionable, context-sensitive governance models. This is especially critical in HE, where governance gaps often emerge between policy ambition and institutional practice. Second, the Dialogue can promote coordinated capacity-building efforts at scale. By encouraging the development of scaffolded training resources as open educational resources—conceived as digital public goods—it can enable equitable access to AI literacy across borders. Such efforts would strengthen not only technical capabilities but also socio-ethical and human rights-based approaches to AI, empowering institutions, educators, and learners to participate meaningfully in governance processes. Third, it can catalyse the development of interoperable governance frameworks by supporting shared tools, benchmarks, and models. For example, promoting institutional AI governance frameworks, national AI–HE coordination mechanisms, and regional initiatives such as a Mediterranean Digital Education Charter would enhance collaboration and mutual learning. These mechanisms can help harmonise ethical standards, data governance practices, and open knowledge ecosystems, while remaining responsive to local realities. Finally, the Dialogue can strengthen digital cooperation by fostering partnerships that support multilingual learning, virtual mobility, and cross-border curriculum development. In doing so, it positions higher education institutions as key actors in advancing responsible AI ecosystems.

Several existing initiatives and partnerships already provide strong foundations for advancing international cooperation on AI governance in higher education, and the Global Dialogue should build explicitly upon these to maximise impact and avoid duplication. A central mechanism is the UNESCO OER Dynamic Coalition, which brings together governments, higher education institutions, and civil society to support the implementation of the UNESCO 2019 OER Recommendation through capacity-building, policy development, and international cooperation . Closely linked is the Dubai Declaration on Open Educational Resources, which provides a recent global roadmap for leveraging AI and digital public goods to enhance equitable access to knowledge, while addressing challenges such as licensing, data protection, and multilingual accessibility . At regional level, networks such as the Union of Universities of the Mediterranean and its Pact for the Mediterranean play a critical role in fostering collaboration across higher education institutions, supporting digital cooperation, mobility, and shared policy development. These initiatives align closely with the vision articulated in Responsible Artificial Intelligence for Digital Cooperation in the Mediterranean Higher Education (Atenas & Stefanelli, 2025), which positions HE as a strategic interface between global frameworks and national implementation, and calls for coordinated, ethical, and open AI integration. The added value of the AI Dialogue would lie in connecting these initiatives into a more coherent global ecosystem. Specifically, it could: Facilitate alignment between regional efforts (e.g. Mediterranean cooperation) and global normative frameworks; Support the scaling of open, interoperable digital public goods, including OER, open data, and shared infrastructures; Promote the co-creation of governance instruments, such as a Mediterranean Digital Education Charter, that harmonise ethical principles and data governance standards; Strengthen pathways from policy to implementation by linking international commitments with institutional governance models in higher education. In this sense, the Dialogue would not replace existing initiatives but amplify their impact—providing a convening space to integrate efforts, share practices, and ensure that AI governance evolves as a coordinated, inclusive, and globally connected endeavour grounded in human rights and open knowledge.

Different stakeholders can contribute to the AI Dialogue through structured, multi-level engagement that reflects their distinct roles while enabling collective governance design. Governments should lead the mapping of national AI strategies and articulate clear national commitments to AI capacity-building, particularly within higher education. This includes aligning national priorities with international frameworks and sharing implementation progress, gaps, and lessons learned. Such mapping exercises would enhance transparency and support policy interoperability across regions. Higher education institutions (HEIs) should act as both contributors and testbeds for governance innovation. Drawing on frameworks such as those highlighted in Responsible Artificial Intelligence for Digital Cooperation in the Mediterranean Higher Education (Atenas & Stefanelli, 2025), HEIs can provide evidence from practice, including institutional AI governance models, AI literacy programmes, and the development of open educational resources (OER) as digital public goods. Their participation is essential to bridge the gap between policy and implementation. International organisations, including UNESCO, can provide normative guidance, convening power, and tools for readiness assessment and peer learning. Existing mechanisms such as the UNESCO OER Dynamic Coalition demonstrate how multi-stakeholder collaboration can support capacity-building, policy development, and knowledge sharing at scale. Civil society, the private sector, and underrepresented groups—including women, persons with disabilities, and Indigenous communities—must be actively included to ensure that governance models are participatory, inclusive, and responsive to societal needs. In terms of format, the AI Dialogue should adopt a hybrid, iterative structure combining: Thematic working groups (e.g. capacity-building, governance interoperability, ethics); Regional labs or observatories to contextualise global principles; Policy mapping sessions to compare national strategies and commitments; Implementation tracks focused on higher education, including roadmaps for AI literacy and governance. The Dialogue can build on frameworks such as the Dubai Declaration on OER, which emphasises international cooperation, capacity-building, and the role of digital public goods in equitable knowledge systems. Ultimately, its added value lies in enabling stakeholders not only to exchange perspectives but to collectively shape inclusive, actionable governance models grounded in human rights, open knowledge, and shared global responsibility.

Global discussions on AI governance continue to underrepresent several key voices, particularly those most directly affected by AI systems in educational and societal contexts. Learners are notably absent from governance processes, despite being primary users of AI-enabled systems in education. Their perspectives on issues such as academic integrity, data privacy, and the pedagogical impact of AI are essential for shaping responsive and legitimate policies. Similarly, academics are often included as technical experts but less frequently as educators who experience the day-to-day implications of AI in teaching, assessment, and curriculum design. Their dual role—as researchers and practitioners—positions them uniquely to inform governance grounded in real educational practice. Persons with disabilities remain insufficiently represented, despite the significant implications of AI for accessibility and inclusion. Without their active participation, AI systems risk reinforcing existing barriers rather than enabling equitable access. Likewise, Indigenous communities and linguistically diverse groups are often marginalised, raising concerns around data sovereignty, cultural representation, and linguistic justice. In addition, open science and open education advocates are frequently overlooked, despite their expertise in building transparent, collaborative, and rights-based knowledge ecosystems. Their work in advancing Open Educational Resources (OER), open data, and open infrastructures is critical for ensuring that AI development aligns with principles of equity, accountability, and public value. Frameworks such as the ROAM-X framework emphasise the importance of multi-stakeholder participation, accessibility, and inclusivity in digital governance. Building on this, inclusion can be strengthened through structured mechanisms such as participatory policy labs, citizen assemblies, and dedicated stakeholder panels within the AI Dialogue. Moreover, capacity-building initiatives—particularly those based on OER as digital public goods—can empower underrepresented groups to engage meaningfully in governance processes. Ensuring accessibility by design, providing multilingual resources, and supporting communities of practice are essential steps. Ultimately, broadening participation is not only a matter of representation but of governance quality, ensuring that AI systems are shaped by diverse perspectives and aligned with human rights and societal needs.

Meaningful and dynamic engagement in the AI Dialogue requires innovative, multi-layered formats that move beyond traditional plenary discussions towards participatory, practice-oriented collaboration. For example, satellite events and regional consultations should be central to the Dialogue's structure. These decentralised formats enable context-sensitive engagement, allowing stakeholders to address regional priorities, map national AI strategies, and articulate capacity-building commitments. They also create space for underrepresented voices—particularly from the Global South—to contribute meaningfully to global governance discussions. Also, the Dialogue should integrate multi-stakeholder co-creation labs or "policy design sprints", where governments, higher education institutions, civil society, and the private sector collaboratively develop governance models, institutional roadmaps, and AI literacy frameworks. These labs could focus on concrete outputs, such as national AI–higher education coordination mechanisms, institutional AI charters, and the development of scaffolded training resources as open educational resources (OER) conceived as digital public goods. Furthermore, thematic knowledge networks and communities of practice should be actively engaged, particularly those advancing open education, inclusive technologies, and ethical AI. Actors such as the UNESCO OER Dynamic Coalition, UNESCO Chairs, and academic networks in open education, technology-enhanced learning, inclusive education, and AI can contribute evidence-based practices, scalable models, and policy expertise. Their involvement ensures that discussions are grounded in existing initiatives and aligned with frameworks such as the UNESCO OER Recommendation and the Dubai Declaration. Also, interactive formats—including participatory workshops, scenario-based simulations, and deliberative citizen panels—can foster deeper engagement by enabling stakeholders to explore real-world challenges and governance trade-offs. These approaches are particularly effective in addressing the socio-ethical implications of AI and promoting inclusive participation. Finally, digital platforms for continuous engagement should complement in-person events, enabling ongoing dialogue, resource sharing, and collaborative drafting of policy recommendations.

Atenas, J., & Stefanelli, C. (2025). Responsible artificial intelligence for digital cooperation in the Mediterranean higher education: Occasional paper. Mediterranean Universities Union (UNIMED). https://doi.org/10.5281/zenodo.15731005 United Nations Educational, Scientific and Cultural Organization. (2024). Dubai Declaration on Open Educational Resources (OER): Digital public goods and emerging technologies for equitable and inclusive access to knowledge (CI/UAI/2024/44/3). World Open Educational Resources Congress (3rd, Dubai, 2024). https://unesdoc.unesco.org/ark:/48223/pf0000392271