Misión Permanente de España ante Naciones Unidas

Spain aligns itself with the written contribution made by the European Union and wishes to make additional comments in national capacity. Spain attaches the highest importance to ensuring that this first edition is substantive, inclusive and sets a credible precedent. A successful first Dialogue should achieve five outcomes. First, it should demonstrate that the United Nations can host a genuinely multistakeholder conversation on AI governance. It should be a space where the scientific community, the private sector, civil society and the technical community participate meaningfully alongside governments. Second, the Dialogue should produce substantive convergence around the four thematic clusters proposed by the Co-Chairs, informed by the first report of the Independent International Scientific Panel. Success means all stakeholders using science as a common baseline. Third, it should deliver an ambitious Co‑Chairs' Summary that consolidates a durable institutional framework for international AI governance. The Summary should reflect the orientations mentioned during the Global Dialogue and grounded in the evidence provided by the Independent Scientific Panel, identify priority areas for cooperation, and set the basis for a light multistakeholder mechanism to sustain momentum between editions. Fourth, the Dialogue should set in motion a credible continuity mechanism. A two-day annual convening cannot sustain the momentum that AI governance demands. Spain has proposed the creation of an AI Governance Action Platform as a light, voluntary, multistakeholder mechanism inspired by the Sevilla Platform for Action (FfD4, 2025). Fifth, the level of political representation will be decisive. Spain encourages participation at Ministerial level or above, as well as at the highest level among private sector, academia and civil society.

• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights
• Interoperability of governance approaches
• Transparency, accountability, and human oversight

1. Social, economic, ethical, cultural, linguistic and technical implications of AI. This cluster hosts some of Spain's most distinctive priorities: multilingualism and linguistic biases as a governance issue; and AI for public services, where Spain, recognised by the OECD as a leader, encourages the Dialogue to share concrete experiences across regions. 2. Protection and promotion of human rights. International human rights law applies fully in the digital sphere. Spain advocates for a human-centric approach as the normative backbone of the Dialogue. Particular attention must be devoted to the protection of children and adolescents. 3. Interoperability of governance approaches. The Dialogue's value lies in fostering coherence, interoperability and mutual understanding across existing initiatives at UN, regional and national level. The proposed AI Governance Action Platform would serve as an operational mechanism to advance this interoperability through voluntary multistakeholder coalitions. 4. Transparency, accountability and human oversight. Without transparency and accountability, safe and trustworthy AI will remain aspirational. The risks of disinformation, misinformation and algorithmic manipulation are eroding trust in institutions and in democracy itself. Spain supports practical policy avenues to address these risks coherently.

To more fully reflect the priorities expressed by a wide range of Member States and UN entities, Spain proposes reinforcing the visibility of several cross-cutting issues that have gained relevance in recent multilateral consultations, within the themes proposed. Information integrity. The expansion of generative AI has heightened international attention on the quality of the information space, content authenticity and institutional resilience. Several UN agencies and Member States have highlighted the need for dedicated discussion on this topic within the Dialogue. AI applications in public services. Health, education, justice and social protection are areas where AI is already being deployed at scale and where there is strong demand for the exchange of experiences, good practices and technical cooperation. Including this dimension would help connect the Dialogue with widely shared development priorities. Protection of children in digital environments. This issue has been underscored by multiple UN actors and could be addressed more explicitly within the human rights related cluster, enabling a coherent and substantive discussion. Spain also underlines that capacity-building and open models are essential for building broad and inclusive consensus, particularly with the Global South. These elements are central to reducing digital divides and enabling equitable participation in the Dialogue. Their inclusion within the 'Bridging AI divides' cluster proposed by the Co-Chairs reflects their structural importance for an inclusive and balanced governance framework. Spain also supports highlighting elements already present in the Co-Chairs' draft: multilingual and culturally inclusive AI, openness and open models, and interoperability between governance approaches.

As an EU Member State, Spain operates within the most comprehensive regulatory framework for AI (the EU AI Act). The governance gaps most relevant to Spain include: the lack of concrete mechanisms to translate international commitments into sustained action between high-level events; the underrepresentation of other languages besides English in the training data, benchmarks and cost structures of major AI models, despite over 500 million native speakers; and growing threats to information integrity from AI-generated synthetic content. Spain also notes that limited access to open and adaptable AI models constrains the ability of public institutions, researchers and smaller actors to audit systems, develop multilingual capabilities and participate meaningfully in global AI governance. This gap disproportionately affects countries with fewer resources and reinforces existing asymmetries. At the same time, Spain sees significant opportunities: AI applied to public services, digital transformation of the economy, and the potential for AI to accelerate progress on the SDGs. Spain is also home to a growing AI research ecosystem, including the Barcelona Supercomputing Center and ELLIS Alicante, and hosts an AI Governance Lab established in the UN headquarter in Valencia. Open and transparent AI models also represent a significant opportunity for Spain, given its scientific capabilities and multilingual environment. They enable more inclusive innovation, strengthen auditability and support the development of AI systems that reflect linguistic and cultural diversity.

The Dialogue can play three distinctive roles. First, as a platform for convergence grounded in science. The report of the Independent International Scientific Panel, whose first in-person meeting was hosted by Spain in Madrid on 22–24 April 2026, should serve as a common baseline, reducing fragmentation and aligning conversations around evidence. Second, as a bridge between initiatives. The Dialogue can foster interoperability and mutual understanding across all the ongoing processes without creating new institutional layers. Third, as a catalyst for sustained action. Spain has proposed an AI Governance Action Platform: a light, voluntary, multistakeholder continuity mechanism inspired by the Sevilla Platform for Action (FfD4, 2025). The Platform would host voluntary coalitions organised around the four thematic clusters, with each initiative co-led by partners from different stakeholder categories. The Co-Chairs' Summary could consolidate clear policy orientations grounded in the Panel's evidence, an inventory of existing initiatives, and a roadmap for the second Dialogue informed by lessons learned.

The Dialogue should actively connect with and build upon existing processes, avoiding duplication and maximising complementarities: - The UN Independent International Scientific Panel on AI, - UNAWG-AI, - The Global Digital Compact and its follow-up mechanisms, including the UN Digital Cooperation Portal and the WSIS review process, - The ITU AI for Good Summit and the WSIS Forum, - The UNESCO Recommendation on the Ethics of AI and its monitoring mechanisms - The Council of Europe Framework Convention on AI, Human Rights, Democracy and the Rule of Law - The Hiroshima AI Process, the OECD AI Principles, and the Global Partnership on AI (GPAI) - AI Safety Institutes and their network - The Internet Governance Forum (IGF) and its Policy Network on AI, - The UN AI Governance for Humanity Lab in Valencia, - National and regional AI governance frameworks, including the EU AI Act. Spain wishes to highlight the availability of the AI Governance for Humanity Lab, as part of the United Nations Office for Digital and Emerging Technologies (ODET) and based in the UN centre in Quart de Poblet (Valencia), as a resource for the whole UN ecosystem. The AI Governance for Humanity Lab focuses on connecting, translating, innovating, and accelerating inclusive, agile, and adaptive cooperation on AI governance. It mobilizes a network of stakeholders across regions and sectors, develops shared operational understanding, and advances practical cooperation tools. In particular, it contributes practice-based insights, operational learning, and collaborative networks that can inform and support the work of UN-led processes - including the Global Dialogue on AI Governance - while remaining distinct from scientific assessment and intergovernmental deliberation functions.

Spain offers the following recommendations on format and structure: High-level participation. Encouraging participation at the highest political and multistakeholder levels would reinforce the visibility and traction of this first edition. Balanced treatment of the four clusters. Ensuring that all four thematic areas receive equivalent attention across the two days would contribute to more coherent participation and discussion. Adequate balance between plenary and interactive formats. The programme should ensure adequate balance between high-level plenary segments and interactive multistakeholder formats, enabling all participants to follow cross-cutting debates while also engaging substantively in thematic discussions. Interactive formats such as policy labs, use-case demonstrations and multistakeholder roundtables would enable more dynamic and results-oriented exchanges. Dedicated space for the Scientific Panel. Beyond the plenary presentation of the Panel's report, Panel members should participate actively in thematic breakout sessions. A segment allowing Member States to react to the Panel's findings would help connect scientific evidence with political orientations. Multilingualism. Spain calls for full respect of all six official UN languages throughout the Dialogue, including in the reporting of outcomes and the accompanying digital platforms.

Developing countries. Many countries of the Global South have not had a meaningful voice in the plurilateral processes that have shaped AI governance to date. The Dialogue's universal character is its greatest asset. Non-English-speaking communities. The dominance of English in AI governance discussions mirrors and reinforces the linguistic biases embedded in AI systems themselves. Children and young people. Those most affected by AI's impact on education, mental health and online safety are largely absent from governance discussions. Their perspectives should be represented through civil society organisations working on children's rights and digital safety. Civil society from the Global South. While Northern civil society organisations are well-represented, their counterparts from the rest of the world face significant barriers.

Structured interaction with the Scientific Panel. Rather than limiting the Panel's contribution to a single plenary presentation, Panel members could participate actively in thematic breakout sessions, offering evidence-based scene-setting and responding to questions from delegations and stakeholders. Multilingual digital platform for year-round inputs. A digital space, hosted by the Joint Secretariat, that maintains a living catalogue of all initiatives under the AI Governance Action Platform, accepts inputs in all six official UN languages, and allows new partners to join existing coalitions throughout the year. Policy labs and use-case demonstrations. Interactive sessions where concrete AI applications in public services, information integrity tools or multilingual AI solutions are demonstrated and discussed by mixed groups of policymakers, technologists and civil society representatives.

The Sevilla Platform for Action (FfD4, 2025). The Sevilla Platform aggregated over 130 voluntary initiatives proposed by coalitions of Member States, international organisations, the private sector and civil society. Spain proposes to transpose this proven model to the AI governance context through the AI Governance Action Platform. AI in Spanish public administration. Spain has been recognised by the OECD as a leader in the use of AI in public services, with applications across tax administration, healthcare, social services and citizen engagement. Spain also highlights the role of ALIA, its public AI infrastructure based on open and adaptable models. ALIA enables transparent evaluation, multilingual fine-tuning and reproducible experimentation, offering a practical pathway for responsible AI development and oversight. By lowering barriers to access and supporting public-interest use cases, ALIA contributes to more inclusive and effective AI governance. Spain is also developing practical governance tools, including the work of the Spanish Agency for AI Supervision (AESIA) on algorithmic assessment methodologies. Spain has also co-funded with UNICEF the largest global study to date on digital childhood, providing a robust evidence base for policies on online safety, wellbeing and age-appropriate design. This initiative demonstrates how international cooperation and large-scale research can support effective and child-centred AI governance. The EU AI Act. As an EU Member State, Spain operates within the most comprehensive risk-based regulatory framework for AI, offering lessons on interoperability and multistakeholder engagement.