EVOLving Leadership

A successful first Global Dialogue on AI Governance would achieve three interrelated outcomes. First, it would produce a clear, actionable, and implementable Co-Chairs' Summary that reflects not only areas of convergence, but also identifies key governance gaps. It should contribute to shaping coherent and forward-looking normative frameworks for AI governance at the global level, beyond high-level principles, including approaches that embed ethical considerations across the full AI lifecycle—from data and model development to deployment and use. Second, the Dialogue would benefit from identifying a small number of cross-cutting issues that connect the four thematic clusters and help integrate technical, societal, and human-rights perspectives. One such emerging area concerns the conditions of human perception, self-representation, and lived experience in AI-mediated environments. As AI systems increasingly shape attention, identity formation, and social cognition, governance frameworks must address not only outputs and impacts, but also the conditions under which human experience is formed. Third, success would include establishing a practical continuity mechanism beyond the inaugural Dialogue. This could take the form of lighter thematic tracks or sustained multi-stakeholder cooperation focused on emerging governance challenges, including human-centered design responsibilities and the protection of human dignity in increasingly immersive digital environments. Together, these outcomes would ensure that the Dialogue not only consolidates existing efforts, but also advances the next horizon of AI governance: frameworks that remain not only safe and trustworthy, but worthy of the human being.

• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Safe, secure and trustworthy AI
• AI capacity-building
• Protection and promotion of human rights

My selection reflects a priority to strengthen AI governance at the intersection of human dignity, societal impact, and system design. First, the protection and promotion of human rights, including transparency, accountability, and human oversight, remains foundational. As AI systems increasingly shape decision-making and social environments, it is essential that governance frameworks not only safeguard rights at the level of outcomes, but also ensure that human agency, autonomy, and dignity are preserved within AI-mediated contexts. Second, the socioeconomic, ethical, cultural, linguistic, and technical implications of AI are of central importance. AI systems are not neutral tools; they shape cultural meaning, identity formation, social cognition, and access to opportunity. Addressing these dimensions is critical for ensuring inclusive and context-sensitive governance, particularly across diverse societies and vulnerable populations. Third, safe, secure, and trustworthy AI is a necessary condition for public confidence and long-term adoption. However, safety must be understood not only in technical terms, but also in relation to broader human impacts, including the integrity of information environments and the conditions under which individuals and communities experience reality. Fourth, AI capacity building is essential to ensure that all regions and stakeholders can meaningfully participate in shaping AI governance. This includes not only technical skills, but also the development of interdisciplinary and cross-cultural competencies needed to address the societal and human dimensions of AI. Together, these priorities support a more integrated approach to AI governance-one that aligns technical robustness with human-centered values and long-term societal resilience.

One important cross-cutting issue that is not yet fully captured is the protection of the integrity of human perception, self-representation, and lived experience in AI-mediated environments. Current AI governance frameworks rightly address safety, transparency, accountability, access, and human rights. However, as AI systems become increasingly embedded across the full lifecycle-from data collection and model design to deployment and interaction-they are not only shaping decisions and outputs, but also the conditions under which human beings perceive, interpret, and experience reality. This emerging domain we term aesthetic rights in the AI era. It concerns the protection of human subjectivity, attention, identity formation, and the dignified representation of persons and communities within digital and algorithmic environments. This issue is inherently cross-cutting. It relates to human rights, as it touches on autonomy, dignity, and freedom of expression; to safety and trust, as perception and meaning-making are increasingly mediated by AI systems; and to socio-cultural implications, as AI shapes collective narratives, cultural visibility, and social cognition. It also has implications across the full AI lifecycle. Governance frameworks may need to consider not only outcomes, but also how data, models, and system design influence perception, attention, and experience. This suggests a need for greater emphasis on human-centered design responsibilities and public awareness, including educational efforts that strengthen individual and societal capacity to navigate AI-mediated environments. Addressing this emerging dimension would help ensure that AI governance protects not only what systems do, but also what remains irreducibly human and what it means to remain human within them.

Across both advanced AI ecosystems and sensitive regional contexts, the governance gaps identified above are already producing uneven and, at times, destabilizing effects. In leading AI development environments, particularly in the United States, rapid technological advancement is outpacing the integration of comprehensive governance frameworks that address not only safety and performance, but also broader societal and human implications. This creates a gap between innovation capacity and the ability to anticipate long-term impacts on perception, identity, and social cohesion. In regions such as the Western Balkans—characterized by complex historical, cultural, and geopolitical dynamics—these gaps can be amplified. Limited institutional capacity, uneven digital literacy, and high exposure to external information ecosystems make societies more vulnerable to manipulation, polarization, and the unintended consequences of AI-mediated environments. A key challenge across both contexts is the absence of governance approaches that address the full AI lifecycle, including how data, models, and system design shape not only decisions, but also human perception, trust, and lived experience. Without this, risks extend beyond technical failures to include erosion of social cohesion and informed agency. At the same time, there is a significant opportunity. Regions that engage early with integrated, human-centered governance approaches—combining technical standards, human rights frameworks, and public capacity building—can strengthen resilience and contribute to more balanced global AI development. Bridging these contexts highlights the importance of governance frameworks that are both globally coherent and locally responsive, ensuring that AI systems support not only innovation, but stable, inclusive, and dignified human development.

The Global AI Dialogue can play a critical role as a bridging mechanism across fragmented governance efforts, enabling greater coherence, trust, and shared direction at the international level. First, it can serve as a platform for alignment between diverse governance approaches emerging across regions, sectors, and institutions. As AI development accelerates, governance responses remain uneven and, at times, disconnected. The Dialogue can help surface areas of convergence while constructively engaging with differences, contributing to more interoperable and mutually reinforcing frameworks. Second, it can elevate cross-cutting issues that are not fully addressed within existing thematic or institutional silos. By integrating technical, societal, and human rights perspectives, the Dialogue can support more holistic governance approaches that reflect the full impact of AI systems across the lifecycle—from design to deployment and use. Third, the Dialogue can strengthen international cooperation by fostering shared understanding and capacity across stakeholders, particularly by connecting advanced AI ecosystems with regions that are more vulnerable to uneven development and external technological influence. This can help reduce fragmentation and support more inclusive participation in shaping AI governance. Finally, the Dialogue can play an important role in shaping forward-looking normative directions. Beyond information exchange, it can contribute to the development of governance approaches that are actionable, adaptable, and capable of informing future international frameworks. In this way, the Dialogue can help move global AI governance from fragmentation toward coordinated, human-centered, and future-oriented cooperation.

The AI Dialogue would benefit from building upon and connecting existing multilateral, multi-stakeholder, and interdisciplinary initiatives, including global foresight and policy networks, that are already addressing different dimensions of AI governance. These include global policy platforms, expert networks, and collaborative foresight initiatives that contribute to understanding long-term implications of AI; regional and international efforts advancing human rights–based approaches to digital governance; and technical and standards-oriented bodies working on safety, interoperability, and responsible system design. In addition, initiatives that integrate societal, cultural, and ethical perspectives into technological development provide an important foundation for more inclusive governance. The added value of the AI Dialogue lies in its ability to connect these otherwise fragmented efforts. Rather than duplicating existing work, it can function as a convergence space that enables knowledge exchange, identifies gaps across domains, and supports the translation of principles into more coherent and implementable governance approaches. In particular, the Dialogue can contribute by: 1. Bridging technical and societal perspectives, ensuring that governance frameworks reflect both system performance and human impact 2. Strengthening links between global and regional efforts, allowing for more context-sensitive implementation 3. Elevating emerging cross-cutting issues that are not yet fully integrated into existing initiatives 4. Supporting continuity through lighter, thematic follow-ups and collaborative tracks. By connecting and amplifying existing work, the AI Dialogue can enhance coherence across the global governance landscape and contribute to more effective, inclusive, and future-ready approaches to AI governance.

Different stakeholders can contribute most effectively if the Dialogue is structured not only around institutional statements, but also around focused, problem-solving exchanges. Governments can bring normative and regulatory perspectives; the technical community can clarify system design and risk; civil society can surface lived impacts; academia and foresight networks can identify emerging issues; and youth, women, Indigenous peoples, persons with disabilities, and communities from the Global South can bring experience-based knowledge often missing from formal governance debates. I would recommend a structure combining high-level plenaries with smaller thematic working sessions, each designed to produce concrete inputs for the Co-Chairs' Summary. These sessions should bring together different stakeholder categories around shared questions, rather than separating them into institutional silos. The Dialogue could also include short "emerging issue" interventions, allowing participants to identify governance gaps not yet fully captured by existing frameworks, such as perception, self-representation, identity, and lived experience in AI-mediated environments. This would help ensure that the Dialogue remains adaptive and future-facing.

Underrepresented voices include women, youth, Indigenous peoples, persons with disabilities, linguistic minorities, artists and cultural workers, educators, and communities from the Global South and smaller states. These groups are often discussed as affected populations, but less often included as knowledge holders and co-designers of governance. They should be included not only through representation quotas, but through meaningful participation formats: preparatory consultations, funded travel support, regional listening sessions, multilingual access, and thematic spaces where lived experience is treated as governance-relevant knowledge. It is especially important to include communities whose identity, visibility, and self-representation are shaped or misrepresented by digital systems. AI governance must account for how technologies affect dignity not only through material outcomes, but also through recognition, voice, perception, and cultural presence.

The Dialogue could benefit from formats that move beyond formal statements toward shared learning and co-creation. These might include intercultural dialogue circles, youth and community labs, scenario-based foresight exercises, artistic and cultural programming, and structured "AI impact hearings" where communities describe how AI-mediated systems affect their lives, identities, opportunities, and rights. Public education should also be treated as part of governance. The Dialogue could support social learning formats that help communities understand AI systems, articulate concerns, and participate in shaping responsible futures. This could include media partnerships, educational toolkits, civic workshops, and even gamified learning models for youth. Most importantly, engagement should cultivate a sense of shared responsibility and interdependence. AI governance is not only a technical or legal task; it is a cultural and civic process through which societies learn how to live with powerful technologies while protecting human dignity, plurality, and the conditions of meaningful participation.

One useful example is the development of Montenegro's national policy framework for interculturalism and social cohesion, which I led within the Directorate for Interculturalism at the Ministry of Human and Minority Rights. This framework positioned interculturalism not only as diversity management, but as a governance approach grounded in human dignity, human rights, social cohesion, and the recognition of profound interconnectedness and interdependence. Its relevance for AI governance lies in the lesson that values-based governance can be institutionalized through public policy, even in smaller and complex societies. Effective AI governance will likewise require more than technical regulation. It will require social architectures that strengthen trust, inclusion, public participation, and shared responsibility. A promising approach would be to combine human-rights-based AI governance with complementary human responsibilities, public education, intercultural dialogue, and participatory policy design. This would help move governance beyond consultation toward meaningful partnership in shaping the conditions of technological development. Such models can support AI governance by creating civic and institutional capacity for social learning, ethical innovation, and human-centered design. They can also help ensure that AI systems serve the public good, protect pluralism, and contribute to a renewed social contract in a world increasingly shaped by intelligent and converging technologies.