Academia Mexicana de Ciberseguridad y Derecho Digital, AMCID

A successful first Global Dialogue on AI Governance should not be measured by the number of principles reaffirmed, but by whether it produces mechanisms capable of reducing fragmentation, managing systemic risk, and enabling practical cooperation across jurisdictions. One meaningful outcome would be the establishment of a shared baseline for AI security and resilience. Today, discussions on "trustworthy AI" often overlook a critical reality: AI systems are increasingly becoming part of critical infrastructure, financial systems, healthcare, public administration, and national security environments. Without cybersecurity, there is no trustworthy AI. The Dialogue should therefore advance a common understanding of AI-related cyber risks — including adversarial attacks, data poisoning, model manipulation, supply chain vulnerabilities, and the misuse of generative AI in influence operations and cybercrime. A second successful outcome would be the creation of operational channels for international cooperation, particularly for developing countries. The AI divide is no longer only about access to models or computing power; it is increasingly about unequal access to security capabilities, technical expertise, incident response capacity, and governance infrastructure. Capacity-building efforts must therefore include AI security, digital resilience, and secure governance frameworks. Third, the Dialogue should avoid becoming exclusively declaratory. It should produce actionable outputs: interoperable risk-management approaches, voluntary technical cooperation mechanisms, repositories of best practices, and pathways for coordinated responses to cross-border AI incidents. Equally important is ensuring that governance discussions remain grounded in reality. The most immediate risks are not hypothetical superintelligence scenarios, but the accelerated industrialization of disinformation, fraud, cyber-enabled coercion, and automated vulnerabilities already affecting societies today. Ultimately, the success of this Dialogue will depend on whether it helps transform AI governance from a fragmented debate into a functional architecture of international trust, security, and accountability

• Safe, secure and trustworthy AI
• AI capacity-building
• Protection and promotion of human rights
• Transparency, accountability, and human oversight

Our organization's priorities are closely aligned with four interconnected areas identified in Resolution 79/325: safe, secure and trustworthy AI; AI capacity-building; protection of human rights; and transparency, accountability, and human oversight. First, we consider safe, secure and trustworthy AI an urgent priority because security remains an underdeveloped dimension in many AI governance discussions. AI systems are increasingly integrated into critical sectors, yet insufficient attention is given to adversarial attacks, model manipulation, supply chain vulnerabilities, synthetic disinformation, and AI-enabled cybercrime. Trustworthy AI cannot exist without resilient and secure digital infrastructures. Second, AI capacity-building is essential, particularly in developing regions. The current AI divide is not limited to access to computing power or models; it also reflects disparities in cybersecurity maturity, regulatory readiness, technical expertise, and institutional capacity. For Latin America and many parts of the Global South, governance without capacity-building risks deepening technological dependence and asymmetries. Third, the protection and promotion of human rights must remain a central pillar of AI governance. The accelerated deployment of AI systems increasingly affects privacy, non-discrimination, due process, freedom of expression, and even cognitive autonomy through emerging neurotechnologies and large-scale behavioral profiling. Human rights safeguards should therefore be embedded into governance frameworks from the design phase, not treated as secondary compliance obligations. Finally, transparency, accountability, and meaningful human oversight are indispensable to democratic legitimacy and public trust. However, transparency alone is insufficient. Governance frameworks must also address explainability limitations, auditability challenges, and responsibility gaps across complex AI supply chains. From our perspective, the most urgent challenge is ensuring that AI governance evolves not only as a framework for innovation, but as a framework for resilience, accountability, and the protection of human dignity in increasingly automated societies.

Yes. While the themes identified in Resolution 79/325 are comprehensive, several cross-cutting and emerging issues deserve more explicit attention due to their growing geopolitical, security, and societal implications. One critical issue is the convergence between AI governance and cybersecurity governance. AI systems are not only tools for innovation; they are increasingly targets, attack surfaces, and operational enablers in cyber conflicts, disinformation campaigns, and critical infrastructure disruptions. The governance debate should therefore address AI security throughout the entire lifecycle of systems, including supply chain integrity, adversarial resilience, incident response coordination, and cross-border risk management. A second emerging issue is the growing challenge of attribution and accountability in AI-enabled operations. Generative AI, autonomous systems, and AI-assisted cyber capabilities are accelerating the industrialization of ambiguity, making it more difficult to distinguish between state action, proxy activity, criminal operations, and automated behavior. This has significant implications for international stability, legal responsibility, and democratic trust. Another important dimension is the concentration of computational power, data, and AI infrastructure. Discussions on AI governance should more directly address digital asymmetries and technological dependency, particularly for developing countries. Governance cannot be sustainable if most states lack meaningful participation in the development, auditing, or security of advanced AI systems. Finally, the Dialogue should increasingly consider the implications of AI in relation to neurotechnologies and cognitive security. Emerging technologies capable of influencing perception, behavior, attention, and neural data introduce unprecedented challenges for privacy, mental autonomy, and human dignity. Ultimately, one of the greatest risks is treating AI governance as a purely technological discussion. In reality, AI governance is increasingly becoming a matter of global security, institutional resilience, economic power, and the protection of democratic societies.

In Latin America, governance gaps in AI are increasingly visible not only in regulatory fragmentation, but also in uneven cybersecurity maturity, limited institutional capacity, and growing technological dependence. These challenges directly affect public institutions, critical infrastructure, financial systems, democratic processes, and citizens' rights. One of the most significant challenges is that AI adoption is accelerating faster than governance and security capabilities. Across the region, organizations are integrating generative AI and automated decision-making systems without sufficiently robust cybersecurity frameworks, risk assessments, audit mechanisms, or incident response capacities. This creates vulnerabilities related to data protection, AI-enabled fraud, disinformation, and attacks against critical services. Another major challenge is the asymmetry between countries developing advanced AI capabilities and those primarily consuming external technologies. Many countries in the Global South remain dependent on foreign infrastructure, models, cloud services, and security architectures, which raises concerns regarding digital sovereignty, strategic dependency, and limited participation in shaping global governance standards. At the same time, the region faces significant capacity gaps in technical expertise, AI security, digital literacy, and interdisciplinary governance. In many sectors, there is still insufficient integration between policymakers, cybersecurity experts, academia, and the technical community. However, these challenges also create important opportunities. Latin America is in a position to contribute a governance perspective strongly grounded in human rights, democratic values, and inclusive digital transformation. The region has an opportunity to help shape governance models that integrate cybersecurity, accountability, and social inclusion from the outset, rather than as corrective measures after harm occurs. There is also growing momentum for regional cooperation, multi-stakeholder collaboration, and capacity-building initiatives. If supported through international cooperation and equitable access to resources and expertise, the region could become an important contributor to more inclusive, resilient, and human-centered AI governance frameworks globally.

The Global Dialogue on AI Governance can play a critical role in reducing the growing fragmentation of AI governance approaches and in building a minimum foundation of international trust, interoperability, and cooperation. At present, AI governance is evolving through parallel national, regional, and sectoral initiatives that often differ significantly in terms of risk classification, regulatory priorities, technical standards, and enforcement mechanisms. Without stronger coordination, these divergences may deepen geopolitical tensions, increase regulatory uncertainty, and create uneven levels of security and protection across countries. The Dialogue can help address this by serving as a neutral multilateral space where governments, the private sector, academia, civil society, and the technical community can exchange not only principles, but also operational experiences, risk-management approaches, and lessons learned from implementation. Importantly, international cooperation should not focus exclusively on innovation and competitiveness. It must also address shared vulnerabilities and systemic risks. AI-related cyber threats, synthetic disinformation, automated fraud, supply chain vulnerabilities, and attacks against critical infrastructure are transnational by nature and cannot be effectively managed through isolated national approaches. The Dialogue can therefore contribute to advancing: -greater interoperability between governance frameworks; -cooperation on AI security and incident response; -shared technical standards and best practices; -coordinated capacity-building efforts for developing countries; -and stronger integration between AI governance, cybersecurity, and human rights protections. Equally important, the Dialogue can help amplify perspectives from regions that are often underrepresented in global technology governance discussions, particularly from the Global South. Inclusive governance is essential for legitimacy and long-term sustainability. Ultimately, the value of the AI Dialogue will depend on whether it evolves beyond a forum for discussion and becomes a platform capable of fostering practical cooperation, reducing asymmetries, and strengthening collective resilience in the face of rapidly evolving technological risks.

The Global Dialogue on AI Governance should build upon existing multilateral, technical, and multi-stakeholder initiatives rather than duplicating efforts. One of its greatest opportunities is to serve as a coordinating platform capable of connecting fragmented governance ecosystems that currently operate in parallel. Several existing frameworks already provide important foundations. The United Nations Educational, Scientific and Cultural Organization Recommendation on the Ethics of Artificial Intelligence has established a globally recognized human rights-based framework for AI governance. In particular, UNESCO's Readiness Assessment Methodology (RAM) offers a valuable operational tool for evaluating national AI governance ecosystems, including legal, institutional, technical, social, and infrastructural capacities. Its methodology is especially relevant for developing countries because it helps identify governance gaps, institutional weaknesses, capacity-building needs, and implementation priorities through a structured and evidence-based approach. In parallel, initiatives such as UNESCO's AI Experts Without Borders network contribute to international cooperation through interdisciplinary expertise, technical assistance, and knowledge-sharing across regions. The Organisation for Economic Co-operation and Development AI Principles and the NIST AI Risk Management Framework have also advanced operational approaches to AI risk management, accountability, and trustworthy AI. Regional initiatives such as the Council of Europe AI Convention process and the European Union AI Act further contribute to discussions on transparency, safety, and rights protections. From a cybersecurity perspective, the Dialogue should also connect with existing cyber norms, resilience frameworks, and incident response cooperation mechanisms. AI governance and cybersecurity governance are increasingly interdependent. The added value of the Global Dialogue lies not in replacing these initiatives, but in fostering interoperability, coordination, and practical cooperation among them while amplifying perspectives from the Global South. Ultimately, the Dialogue can help move global AI governance from fragmented principles toward more coherent, operational, and implementable international cooperation.

The effectiveness of the Global Dialogue on AI Governance will depend largely on whether it enables meaningful participation across disciplines, sectors, and regions, while moving beyond purely declaratory discussions toward practical cooperation. Different stakeholders contribute distinct but complementary perspectives. Governments play a critical role in advancing public policy, regulatory coordination, and international norms. The private sector contributes technical expertise, operational experience, and insight into emerging risks and real-world deployment challenges. Academia and the technical community are essential for independent research, standards development, auditing methodologies, and evidence-based policy analysis. Civil society organizations help ensure that governance frameworks remain aligned with human rights, inclusion, transparency, and public accountability. International organizations can facilitate coordination, capacity-building, and interoperability across governance approaches. From our perspective, one of the most important priorities is ensuring stronger integration between technical and policy communities. AI governance discussions often remain siloed, while many risks — particularly cybersecurity risks — require interdisciplinary collaboration. In terms of format and structure, the Dialogue would benefit from combining high-level policy discussions with more operational and interactive formats. For example: multi-stakeholder roundtables focused on specific risks or sectors; technical-policy workshops; scenario-based exercises on AI-related incidents; regional consultations reflecting different governance realities; and implementation-focused sessions translating principles into actionable practices. The Dialogue should also ensure meaningful participation from developing countries and underrepresented regions, not only as observers but as active contributors to agenda-setting and governance design. Importantly, the Scientific Panel's findings should not remain purely academic outputs. Dedicated sessions should focus on translating scientific and technical assessments into implementable governance recommendations, particularly regarding cybersecurity, accountability, risk management, and capacity-building. Ultimately, the Dialogue's success will depend on whether it creates an environment where cooperation becomes operational, inclusive, and capable of responding to rapidly evolving technological and geopolitical realities.

Several voices remain significantly underrepresented in global AI governance discussions, particularly those from the Global South, Indigenous communities, small and developing states, local technical communities, and interdisciplinary experts working at the intersection of cybersecurity, human rights, law, and emerging technologies. In many cases, global governance debates continue to be shaped primarily by countries and corporations with advanced technological and computational capabilities. This creates an imbalance where regions that are disproportionately affected by technological dependency, cyber vulnerabilities, disinformation, and capacity gaps often have limited influence over the standards, norms, and governance models being developed. Latin America, Africa, and other developing regions frequently face structural barriers to participation, including limited access to funding, technical infrastructure, policy networks, language accessibility, and institutional capacity. As a result, governance discussions risk overlooking the realities of countries that are primarily AI adopters rather than AI developers. Indigenous and culturally diverse communities are also often excluded, despite the significant implications AI systems may have for linguistic diversity, cultural heritage, collective rights, and data governance. Equally underrepresented are cybersecurity practitioners and experts in digital resilience, even though security vulnerabilities increasingly shape the societal risks associated with AI deployment. To address these gaps, the Dialogue should prioritize inclusive participation mechanisms beyond formal representation. This could include: regional preparatory consultations; multilingual participation formats and documentation; financial and institutional support for underrepresented stakeholders; stronger inclusion of technical and civil society communities from developing regions; and interdisciplinary participation that integrates cybersecurity, digital rights, ethics, law, and governance expertise. Importantly, inclusion should not be symbolic. Underrepresented actors must have meaningful opportunities to shape agendas, contribute to decision-making processes, and influence policy outcomes. Ultimately, global AI governance cannot achieve legitimacy or long-term effectiveness if large parts of the world remain structurally excluded from defining the rules that will shape increasingly automated societies.

To foster meaningful and dynamic engagement, the Global Dialogue on AI Governance should move beyond traditional plenary formats and incorporate more operational, interdisciplinary, and problem-oriented engagement mechanisms. One effective approach would be the use of scenario-based simulations and tabletop exercises focused on real-world AI governance challenges. For example, participants could examine coordinated responses to AI-enabled cyber incidents, synthetic disinformation campaigns, failures in automated decision-making systems, or cross-border impacts affecting critical infrastructure. These formats encourage practical cooperation and help bridge the gap between abstract principles and implementation realities. The Dialogue could also benefit from multi-stakeholder policy labs that bring together governments, industry, academia, cybersecurity practitioners, technical experts, and civil society to collaboratively develop governance approaches for specific risks or sectors. These smaller, solution-oriented environments are often more productive than exclusively formal interventions. Regional and cross-regional dialogue tracks would also help reflect different governance realities and capacity gaps, particularly for developing countries and underrepresented regions. Importantly, meaningful inclusion should not depend solely on the ability to travel or secure institutional funding. Many experts, civil society representatives, technical communities, researchers, and grassroots organizations — particularly from the Global South — remain excluded from global governance discussions due to financial and logistical barriers. The Dialogue should therefore adopt hybrid participation models with equal opportunities for remote engagement, including interactive virtual roundtables, remote intervention rights, multilingual participation tools, digital collaboration platforms, and accessible asynchronous contribution mechanisms. Dedicated fellowship, sponsorship, or travel-support programs should also be considered to ensure more equitable representation. Without this, global AI governance risks reproducing existing technological and geopolitical asymmetries. Additionally, interactive sessions integrating the findings of the Independent International Scientific Panel could help translate technical assessments into actionable governance recommendations. Ultimately, the most effective formats will be those capable of fostering genuine interdisciplinary exchange, practical cooperation, and inclusive participation across regions, sectors, and levels of technical capacity.

Several existing policies, frameworks, and governance approaches already provide valuable foundations for more effective and implementable AI governance. At the international level, the United Nations Educational, Scientific and Cultural Organization Recommendation on the Ethics of Artificial Intelligence represents one of the most globally inclusive normative frameworks, particularly because it integrates human rights, governance, sustainability, and capacity-building considerations. Its Readiness Assessment Methodology (RAM) is especially valuable as a practical tool for helping countries identify governance gaps, institutional capacities, regulatory needs, and implementation priorities in a structured and evidence-based manner. The NIST AI Risk Management Framework also offers an important operational model by translating high-level governance principles into risk-management practices focused on reliability, resilience, accountability, and security throughout the AI lifecycle. Its flexibility makes it particularly useful across sectors and governance environments. From a regional perspective, the European Union AI Act provides an important example of a risk-based governance approach, particularly regarding high-risk systems, transparency obligations, and accountability mechanisms. While implementation challenges remain, it contributes significantly to operationalizing AI governance discussions. In the cybersecurity domain, existing incident response coordination mechanisms, threat intelligence-sharing models, and critical infrastructure resilience frameworks also provide relevant lessons for AI governance. AI systems increasingly operate within interconnected digital ecosystems, meaning governance approaches should integrate cybersecurity, supply chain security, and resilience-by-design principles from the outset. Equally important are multi-stakeholder and interdisciplinary approaches. Initiatives such as UNESCO's AI Experts Without Borders network demonstrate the value of collaborative expertise, capacity-building, and cross-regional cooperation. As Women 4 Ethical AI as well Ultimately, effective AI governance will require combining normative principles with operational tools, technical standards, interdisciplinary collaboration, and inclusive participation mechanisms capable of adapting to rapidly evolving technological and geopolitical realities.