WHO Europe

success would mean moving from a symbolic launch to a credible, ongoing process with clear priorities, broad participation, and follow-through after Geneva. The strongest outcome would be practical convergence around a few issues where cooperation is possible now: safety and trustworthiness, human rights, transparency, accountability, and reducing digital divides/

• Protection and promotion of human rights
• Transparency, accountability, and human oversight
• Interoperability of governance approaches
• Safe, secure and trustworthy AI

These four give the strongest governance-oriented package because they combine normative safeguards, operational oversight, system safety, and cross-border policy coherence. They also fit a rights-based and policy-oriented approach better than broader themes, because they are the areas most likely to produce concrete safeguards, accountability mechanisms, and practical international coordination.

Market concentration and compute access. Global AI governance should address who controls compute, chips, cloud infrastructure, and foundation model ecosystems, because concentration can undermine both fair competition and meaningful participation by developing countries. Also, remedy, redress, and incident reporting. Transparency and accountability are important, but affected individuals and institutions also need practical channels for reporting harms, independent review, and access to remedy when AI systems fail or discriminate.

WHO/Europe has explicitly identified an urgent shortage of trained professionals who can manage public health informatics, data science, and integrated electronic data systems, which directly limits safe adoption of more advanced AI-enabled health tools. Financing is another constraint, because several countries in the eastern part of the region still rely heavily on donor-funded or fragmented development models rather than stable long-term funding for digital health operations and governance. At the same time, EECA has strong opportunities. The region is already building foundations through national digital health strategies, telehealth expansion, electronic reporting, immunization registries, and stronger data-use capacities for surveillance and public health response. There is also real opportunity to align around common standards, interoperability, and human-centered governance so that AI in health improves access, continuity of care, and emergency preparedness rather than deepening inequality.

The AI Dialogue can play a valuable role by giving the world a standing UN platform where all countries, not only major AI powers, can shape norms, share experience, and coordinate responses to cross-border risks. First, it can help build a shared baseline on priority issues such as safe, secure and trustworthy AI, interoperability of governance approaches, capacity gaps, and the broader socioeconomic implications of AI. Second, it can reduce duplication by serving as a convening space that complements work already happening in bodies such as the OECD, G7, and regional organizations rather than competing with them. Third, it can elevate countries that are often rule-takers in AI governance by ensuring that developing countries have a voice in agenda-setting, not just in implementation.

The most relevant foundations are the OECD AI Recommendation and the OECD-GPAI partnership, the Global Digital Compact and its scientific panel track, and ongoing work in the G7, IGF, and regional organizations.

Governments = should contribute national experience, regulatory lessons, and concrete capacity needs, especially on interoperability and implementation. UN agencies, regional organizations, and standards bodies = should help map existing initiatives and reduce duplication across forums. The scientific community and the UN scientific panel = should provide independent evidence, risk assessments, and emerging-issue briefings to anchor the dialogue in facts rather than politics alone. Civil society, academia, and affected communities = should highlight rights impacts, accountability gaps, and local realities that may otherwise be missed, particularly from the Global South. Private-sector actors = should contribute technical knowledge, safety practices, and implementation experience, but within clear transparency and public-interest guardrails.

Global AI governance still underrepresents Global South governments and researchers, grassroots civil society, Indigenous peoples, people with disabilities, workers likely to be affected by automation, and communities whose languages and data are poorly reflected in dominant AI systems

Regional pre-dialogues = that feed into the global meeting, so countries and stakeholders arrive with clearer priorities and the main session is not overloaded with first-round interventions. Thematic policy labs = where governments, researchers, civil society, and industry work on one concrete issue, such as interoperability, capacity-building, or accountability, and produce short option papers. Scenario-based exercises = that test how different governance models would respond to real cross-border AI risks, which would make discussions more practical and evidence-based. Fishbowl or listening sessions = reserved for underrepresented voices, including Global South actors, affected communities, and smaller states, so participation is not dominated by the most powerful delegations. Implementation clinics = where countries and institutions share specific governance challenges, lessons, and requests for support, linking dialogue to capacity-building rather than only norm-setting.

The OECD AI Principles = remain one of the clearest international policy baselines because they link innovation to human rights, robustness, transparency, accountability, and inclusive growth. UNESCO's Recommendation on the Ethics of AI = is especially useful because it is global in scope and is backed by implementation tools rather than principles alone The EU AI Act = offers an important regulatory model by imposing binding obligations on high-risk AI systems, including requirements related to risk management, data governance, transparency, human oversight, and post-market monitoring