Blig Consulting / ISOC Puerto Rico

A successful first Global Dialogue on AI Governance should produce practical, inclusive, and actionable outcomes that help governments, civil society, academia, the technical community, and the private sector move from high-level principles to responsible implementation. First, the Dialogue should establish a shared understanding that AI governance is not only a regulatory issue. It is also a matter of cybersecurity, data governance, human rights, risk management, institutional accountability, and digital resilience. Second, it should identify priority areas where international cooperation is urgently needed, particularly for developing regions, small island jurisdictions, and underserved communities. The AI divide should be addressed as a multidimensional issue involving infrastructure, connectivity, digital literacy, cybersecurity readiness, data quality, access to computing resources, and local technical capacity. Third, the Dialogue should promote interoperability among AI governance frameworks to avoid fragmented, duplicative, or conflicting requirements. Organizations need practical guidance that can align with existing cybersecurity, privacy, data protection, procurement, and risk management practices. Fourth, the Dialogue should reinforce the need for meaningful human oversight in high-impact AI use cases, especially in public services, financial services, education, healthcare, employment, cybersecurity operations, and access to essential rights or benefits. Finally, the Dialogue should result in concrete next steps, including governance templates, risk assessment tools, capacity-building programs, implementation playbooks, and mechanisms for continued multistakeholder participation. Its success should be measured not only by policy alignment, but by whether smaller jurisdictions and organizations can realistically adopt safer, more trustworthy, and human-centered AI practices.

• AI capacity-building
• Safe, secure and trustworthy AI
• Interoperability of governance approaches
• Transparency, accountability, and human oversight

I selected these priorities because they are essential for translating AI governance principles into responsible and practical implementation. Capacity gaps must be addressed first because many organizations and jurisdictions are adopting AI faster than their governance, cybersecurity, data management, and workforce capabilities can mature. For Puerto Rico, the Caribbean, and other developing or smaller jurisdictions, the AI divide is not only about access to AI tools. It also includes connectivity, infrastructure, local expertise, language inclusion, cloud readiness, cybersecurity maturity, and institutional capacity. Safe, secure, and trustworthy AI systems are also a priority because AI creates new operational, ethical, legal, and cybersecurity risks. These include data leakage, model manipulation, adversarial inputs, prompt injection, insecure integrations, deepfake-enabled fraud, misinformation, and overreliance on automated outputs. Trustworthy AI requires secure design, responsible deployment, monitoring, accountability, and incident response. Interoperability of governance approaches is critical because organizations already face multiple frameworks and requirements related to privacy, cybersecurity, data protection, resilience, procurement, and risk management. AI governance should complement these existing structures rather than create fragmented or conflicting obligations. Transparency, accountability, and human oversight are necessary for high-impact AI use cases. AI systems should support human decision-making, not replace institutional responsibility. In areas such as public services, finance, education, healthcare, employment, cybersecurity, and access to benefits or rights, human oversight should be proportionate to the level of risk and potential harm. Together, these four priorities create a balanced approach: inclusive capacity-building, secure and trustworthy systems, coherent governance, and human-centered accountability.

Yes. A key cross-cutting issue that should receive stronger attention is the convergence of AI governance, cybersecurity, cyber resilience, and third-party technology risk. AI systems do not operate in isolation. They depend on data pipelines, cloud platforms, APIs, identity systems, software supply chains, vendors, models, and human users. Weaknesses in any of these areas can undermine the safety, reliability, and trustworthiness of AI systems. Therefore, AI governance should explicitly address cybersecurity controls, secure architecture, access management, data classification, vendor due diligence, incident response, auditability, and resilience planning. Another emerging issue is the misuse of generative AI for deception, fraud, impersonation, social engineering, deepfakes, and disinformation. These risks can affect elections, financial institutions, public trust, crisis response, education, journalism, and vulnerable communities. The Global Dialogue should encourage practical safeguards, awareness programs, detection capabilities, and response mechanisms for AI-enabled manipulation and fraud. A third cross-cutting issue is responsible AI procurement. Many organizations, especially in the public sector and smaller jurisdictions, will adopt AI through third-party platforms rather than building their own systems. Procurement processes should include requirements for transparency, data protection, security, model governance, explainability, human oversight, contractual accountability, data retention, and exit strategies. Finally, the Dialogue should consider the needs of small island jurisdictions and regions exposed to natural disasters, infrastructure constraints, and limited technical resources. AI governance should be scalable, realistic, and implementable for organizations with different levels of maturity. Otherwise, global AI governance may unintentionally widen the gap between well-resourced institutions and those still developing their digital and cybersecurity foundations.

Puerto Rico, the Caribbean, and the broader Latin America and Caribbean region are being affected by AI governance gaps in a very practical way. AI adoption is accelerating across education, government, financial services, professional services, cybersecurity, and small businesses, but governance maturity is not advancing at the same pace. Many organizations are using AI tools without fully mature policies for data protection, acceptable use, human oversight, vendor risk management, cybersecurity controls, or accountability. One of the most significant challenges is the AI capacity gap. Smaller jurisdictions and organizations often face limitations in infrastructure, specialized talent, AI literacy, cybersecurity maturity, data governance, and access to trusted technical guidance. This increases the risk of unsafe adoption, especially when AI tools are used with sensitive information or in high-impact decision-making environments. Another challenge is the growing cybersecurity and fraud risk associated with AI. Generative AI can be misused for phishing, impersonation, deepfakes, misinformation, social engineering, and automated fraud. At the same time, organizations are integrating AI through cloud platforms, APIs, third-party tools, and internal workflows, creating new risks related to data leakage, insecure configurations, and lack of monitoring. However, the opportunities are significant. AI can strengthen public services, education, cybersecurity operations, fraud detection, business productivity, disaster response, and digital inclusion. For the Caribbean, AI can also support resilience, workforce development, and modernization of public and private sector services. The key opportunity is to develop practical, risk-based, and interoperable governance models that smaller jurisdictions can realistically implement. This should include AI literacy, cybersecurity safeguards, responsible procurement, data governance, human oversight, and capacity-building tailored to local realities.

The AI Dialogue can play a critical role as a neutral, inclusive, and practical platform for advancing international cooperation on AI governance. Its value should be to connect governments, civil society, academia, the technical community, international organizations, and the private sector around shared priorities without creating unnecessary fragmentation. First, the Dialogue can help develop a common understanding of what responsible AI governance requires across different regions, sectors, and levels of maturity. This includes alignment on safe, secure, and trustworthy AI; transparency; accountability; human oversight; cybersecurity; data governance; and respect for human rights. Second, the Dialogue can support interoperability among governance approaches. Many countries and organizations are already adopting AI-related policies, standards, laws, and frameworks. The Dialogue can help identify common principles, reduce duplication, and encourage practical alignment among existing initiatives. Third, the Dialogue can elevate the needs of developing regions, small island jurisdictions, and underrepresented communities. International cooperation should not be shaped only by large economies or large technology providers. It should also reflect the realities of regions facing capacity gaps, infrastructure limitations, cybersecurity challenges, language barriers, and limited access to technical expertise. Fourth, the Dialogue can promote practical capacity-building. This should include AI literacy, governance templates, risk assessment tools, responsible procurement guidance, cybersecurity safeguards, and implementation playbooks that smaller institutions can realistically adopt. Finally, the Dialogue can serve as a continuous mechanism for trust-building and shared learning. AI governance will continue to evolve, so international cooperation must be sustained, adaptive, and based on real-world lessons, not only high-level policy statements.

The AI Dialogue should build upon and connect with existing international, regional, technical, and standards-based initiatives rather than duplicate them. Relevant mechanisms include the United Nations system, UNESCO's work on AI ethics, the International Telecommunication Union's AI for Good ecosystem, the Internet Governance Forum, the OECD AI Principles, regional digital transformation initiatives, and recognized risk-based frameworks such as the NIST AI Risk Management Framework. It should also connect with cybersecurity, privacy, data protection, and digital resilience communities. AI governance cannot be separated from cybersecurity governance, third-party risk management, data classification, cloud security, incident response, and digital trust. Existing cybersecurity frameworks and technical communities can provide practical experience on how to govern complex technologies in operational environments. The Dialogue should also engage academia, civil society, professional associations, Internet technical community groups, and local or regional organizations working on digital inclusion and capacity-building. These actors are essential to ensure that AI governance is not limited to government-to-government or large industry discussions. The added value of the AI Dialogue is that it can serve as a global coordination layer. It can identify areas of convergence among existing initiatives, highlight gaps, elevate the needs of developing regions and small island jurisdictions, and translate high-level principles into practical guidance. For Puerto Rico and the Caribbean, the Dialogue can add particular value by recognizing the importance of capacity-building, resilience, language inclusion, cybersecurity readiness, and practical governance models for smaller jurisdictions. Its contribution should be to connect existing work, reduce fragmentation, and help make AI governance more inclusive, interoperable, secure, and implementable.

Different stakeholders should contribute based on their respective roles, expertise, and lived realities. Governments can provide policy direction, legal frameworks, public sector priorities, and international cooperation mechanisms. The private sector can contribute technical innovation, implementation experience, responsible AI practices, and lessons learned from deploying AI systems at scale. Academia can provide research, evidence-based analysis, independent evaluation, and education. Civil society can represent public interest, human rights, inclusion, accountability, and the perspectives of affected communities. The technical community can contribute expertise on Internet infrastructure, cybersecurity, interoperability, standards, data governance, and operational risks. The AI Dialogue should be structured to avoid being limited to formal statements. It should include plenary sessions, thematic breakouts, technical roundtables, regional consultations, and practical case-study discussions. Each session should include balanced participation from governments, civil society, academia, private sector, and technical experts. The format should also allow written submissions, remote participation, multilingual access, and structured opportunities for small jurisdictions and underrepresented communities to contribute. This is important for regions such as Puerto Rico and the Caribbean, where AI governance challenges are closely connected to digital inclusion, cybersecurity readiness, public sector modernization, resilience, and capacity-building. The Dialogue should produce practical outputs, not only policy summaries. These could include governance templates, risk assessment tools, responsible procurement guidance, capacity-building priorities, cybersecurity safeguards, and implementation playbooks. A successful structure should combine high-level policy dialogue with operationally useful recommendations that stakeholders can apply in real-world environments.

Several voices remain underrepresented in global AI governance discussions. These include small island jurisdictions, Caribbean communities, developing regions, small and medium-sized enterprises, local governments, public sector practitioners, educators, cybersecurity professionals, technical community groups, and communities directly affected by limited digital infrastructure or limited access to specialized technical expertise. In many AI governance discussions, the perspectives of large economies, major technology companies, and highly resourced institutions tend to dominate. While those perspectives are important, they do not fully reflect the realities of smaller jurisdictions or organizations that must adopt AI with limited budgets, limited staffing, immature data governance, cybersecurity gaps, language barriers, and infrastructure constraints. Puerto Rico and the Caribbean should be more meaningfully included because the region brings a valuable perspective on resilience, digital inclusion, disaster response, education, public sector modernization, and cybersecurity capacity-building. These issues are directly relevant to responsible and inclusive AI adoption. Underrepresented communities could be included through regional consultations, remote participation, multilingual engagement, targeted invitations, travel support, capacity-building workshops, and dedicated sessions for small island developing states, Caribbean stakeholders, and local technical communities. The Dialogue should also create mechanisms for written input before and after the event, so participation is not limited to those able to attend in person. Inclusion should not be symbolic. It should influence the agenda, the final outcomes, and the practical tools developed after the Dialogue. AI governance will be more legitimate and effective if it reflects the realities of communities with different levels of digital maturity, infrastructure readiness, institutional capacity, and exposure to risk.

The AI Dialogue should use engagement formats that go beyond traditional panels and formal speeches. To foster meaningful participation, the Dialogue should combine high-level discussions with practical, interactive, and problem-solving formats. One useful format would be scenario-based governance workshops. Participants could analyze realistic AI use cases involving public services, financial services, education, healthcare, cybersecurity, disaster response, misinformation, or public procurement. These scenarios would help identify practical risks, governance gaps, human oversight requirements, and implementation safeguards. A second format would be multistakeholder roundtables organized by theme and region. These should include governments, civil society, academia, the private sector, and the technical community. Small-group discussions can generate more substantive input than large plenary sessions alone. A third format would be a "governance clinic" model, where jurisdictions or organizations present specific AI governance challenges and receive structured feedback from experts in cybersecurity, law, ethics, data governance, procurement, and public policy. The Dialogue could also include regional listening sessions, virtual participation channels, multilingual written submissions, and digital collaboration spaces before and after the event. This would allow stakeholders from smaller jurisdictions and developing regions to contribute even if they cannot attend in person. Another valuable format would be a practical deliverables lab focused on producing usable outputs such as model AI policies, risk assessment templates, responsible procurement checklists, cybersecurity control mappings, and human oversight guidelines. The most effective engagement model would be one that connects policy, technical expertise, and real-world implementation. The goal should not only be dialogue, but the creation of practical tools and partnerships that help organizations govern AI responsibly, securely, and inclusively.

Effective AI governance should combine policy, risk management, cybersecurity, procurement, education, and continuous oversight. Useful examples include risk-based frameworks such as the NIST AI Risk Management Framework, ISO/IEC 42001 for AI management systems, the OECD AI Principles, and UNESCO's work on AI ethics. These approaches provide a foundation for trustworthy, human-centered, and accountable AI governance. At the organizational level, good practices should include an AI acceptable use policy, an AI system inventory, a risk-tiering methodology, data classification requirements, and an AI impact assessment process. These tools help organizations understand where AI is being used, what data is involved, what risks exist, and what level of oversight is required. Another effective practice is responsible AI procurement. Organizations should require vendors to disclose how AI systems use data, whether data is retained or used for model training, what security controls are in place, how human oversight is supported, and what contractual accountability exists. This is especially important for public sector entities and smaller organizations that adopt AI through third-party platforms. Cybersecurity should also be embedded into AI governance. This includes secure architecture, identity and access management, API security, logging, monitoring, incident response, model abuse prevention, and third-party risk management. AI governance should align with existing cybersecurity and resilience programs rather than operate separately. Finally, practical capacity-building is essential. AI literacy programs, role-based training, governance playbooks, human oversight guidelines, and sector-specific use case templates can help organizations adopt AI responsibly. For Puerto Rico, the Caribbean, and similar jurisdictions, scalable and practical tools are especially important because effective governance must be realistic for institutions with different levels of maturity, resources, and technical capacity.