Universidad Continental

A successful first Global Dialogue on AI Governance—particularly under WHO leadership—should move beyond principles toward actionable alignment, implementation pathways, and measurable commitments. First, success would mean achieving consensus on a globally adaptable governance framework that integrates ethics, regulation, and interoperability. This framework should not remain aspirational, but translate into a "minimum viable governance architecture" that countries—especially low- and middle-income countries (LMICs)—can operationalize. As highlighted in the "Digital Health House" approach, governance must anchor infrastructure and data systems to avoid fragmentation and ensure scalability . Second, a key outcome would be clear commitments on capacity building as a central pillar of AI governance. Evidence from Latin America demonstrates that the primary bottleneck is not technology, but the shortage of trained human capital capable of implementing and governing AI systems effectively . The Dialogue should therefore result in concrete agreements to institutionalize digital health and AI competencies in health and public sector education systems, alongside the creation of regional training hubs. Third, success would require embedding equity-by-design and socio-technical approaches into global AI governance. AI in health cannot be reduced to algorithms alone; it must account for cultural, organizational, and systemic realities. As shown in applied AI for tuberculosis diagnosis, technological solutions only succeed when aligned with socio-technical and contextual factors. This implies commitments to transparency, bias mitigation, and inclusive co-design with communities. Finally, the Dialogue should produce a roadmap with accountability mechanisms, including indicators, financing pathways, and country-level implementation pilots aligned with the next Global Digital Health Strategy (2028–2033). In essence, success is not a declaration—it is a shift: from fragmented innovation to governed, equitable, and workforce-enabled AI systems at scale.

• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI

My selection reflects a core conviction: the success of AI governance-particularly in health-depends less on technological sophistication and more on human capacity, contextual adaptation, and ethical alignment. First, AI capacity-building is the most urgent priority. Across low- and middle-income countries, including in Latin America, the principal barrier is not access to AI tools but the shortage of professionals capable of designing, implementing, regulating, and evaluating them. Without sustained investment in workforce development-integrating AI, digital health, and data governance into health and public sector education-AI risks remaining fragmented, externally driven, and unsustainable. Capacity-building must therefore move from short-term training to institutionalized, system-level competencies, supported by regional hubs and South-South collaboration. Second, the social, economic, ethical, cultural, linguistic, and technical implications of AI are inseparable from its real-world impact. Evidence from AI applications in health demonstrates that outcomes are shaped not only by algorithms, but by socio-technical contexts, including cultural norms, organizational readiness, and structural inequities. If these dimensions are not explicitly addressed, AI may inadvertently reinforce disparities, introduce bias, or fail to be adopted. From my perspective, these two priorities are deeply interconnected. Capacity-building must explicitly incorporate ethical reasoning, cultural competence, and contextual adaptation, ensuring that AI systems are trustworthy, inclusive, and aligned with local needs. This includes promoting transparency, explainability, and community engagement in the design and deployment of AI solutions. Ultimately, my engagement focuses on advancing a people-centered, equity-driven approach to AI governance, where countries are not passive adopters of technology, but active co-creators of solutions that strengthen their own systems and improve population outcomes.

Socio-technical system integration should be elevated as a distinct priority. Evidence shows that AI performance and impact depend not only on algorithms but on how they interact with workflows, institutions, and cultural contexts. Without this perspective, even technically robust solutions may fail or generate unintended consequences.

In Peru and across Latin America, gaps in AI governance are already shaping both the trajectory and equity of digital transformation, particularly in health systems. The most significant challenge is fragmentation. While AI initiatives are growing, they are often implemented as isolated pilots without alignment to national strategies, standards, or regulatory frameworks. This reflects broader governance gaps in interoperability, data standards, and institutional coordination, limiting scalability and long-term impact. Closely linked is the deficit in human capital: there is a critical shortage of professionals trained not only in AI development, but in its governance, evaluation, and ethical oversight. As a result, countries risk becoming dependent on external technologies, with limited capacity to adapt them to local needs. Another major challenge is data readiness. Many systems lack high-quality, interoperable, and representative datasets, which undermines the reliability, fairness, and safety of AI applications. This is compounded by insufficient data governance frameworks, raising concerns around privacy, bias, and accountability. At the same time, these gaps create important opportunities. First, there is a unique window to design "leapfrog" governance models—building integrated, standards-based architectures from the outset, rather than retrofitting legacy systems. Second, the region can position itself as a leader in equity-driven, people-centered AI, embedding socio-technical and cultural considerations into governance frameworks. Third, investments in capacity-building and regional collaboration (including South–South networks) can accelerate the development of a skilled workforce and shared infrastructure. Ultimately, the impact of current governance gaps is a double-edged sword: they constrain progress today, but—if addressed strategically—offer a pathway to more inclusive, resilient, and context-aware AI ecosystems in the future.

The WHO Global Dialogue on AI Governance can play a catalytic role by shifting international cooperation from fragmented discussions to **coordinated, implementation-oriented action**, particularly for health and public sector systems. First, the Dialogue can serve as a **neutral convening platform** that aligns Member States, academia, industry, and multilateral partners around a shared governance vision. This is especially important to bridge asymmetries between high-income and low- and middle-income countries (LMICs), ensuring that global norms are not only defined by a few actors but reflect diverse system realities. Second, it can advance the development of **globally interoperable governance frameworks**, including common principles, standards, and reference architectures that countries can adapt locally. Harmonization is essential to address cross-border challenges such as data flows, AI safety, and regulatory coherence, while avoiding duplication and policy fragmentation. Third, the Dialogue can promote **collective capacity-building mechanisms**, including regional training hubs, knowledge-sharing platforms, and South–South and triangular cooperation. Strengthening human capital is the foundation for effective AI governance and for enabling countries to move from passive adoption to active co-creation. Fourth, it can facilitate **shared infrastructure and public goods**, such as open standards, curated datasets, evaluation frameworks, and benchmarking tools, particularly for priority areas like health. This would reduce barriers to entry and accelerate responsible innovation. Finally, the Dialogue can help establish **accountability and follow-up mechanisms**, including measurable indicators, country commitments, and pilot implementations aligned with the next Global Digital Health Strategy (2028–2033). In essence, its greatest value lies in transforming global dialogue into **collective capacity, coordinated governance, and equitable participation**, ensuring that AI advances as a global public good rather than a source of widening disparities.

The AI Dialogue should build upon and connect with a range of existing global, regional, and sector-specific initiatives that are already advancing responsible AI, particularly in health. At the global level, key foundations include the WHO Global Strategy on Digital Health (2020–2025) and its extension, the UNESCO Recommendation on the Ethics of Artificial Intelligence, and ongoing efforts under the Global Digital Compact. These initiatives provide normative frameworks on ethics, governance, and equity. The Dialogue can add value by translating these principles into implementable, country-level roadmaps, with measurable indicators and operational guidance. At the regional and sectoral level, initiatives such as PAHO's digital health agenda, national AI strategies, and academic–government partnerships are critical. In Latin America, there is growing momentum around capacity-building networks, digital health architectures, and interoperability frameworks, but these remain unevenly developed. The Dialogue can serve as a platform to harmonize efforts, promote South–South collaboration, and scale successful models across countries. Importantly, the Dialogue should also recognize and build upon emerging bottom-up initiatives, such as the Arequipa Manifesto 2025 on the Ethical Use of AI in Medical Education, which emphasizes a person-centered approach, ethical training, data protection, and equitable access to AI . This manifesto reflects a growing consensus from academia and health professionals that AI must complement—not replace—clinical judgment and human values, while being supported by robust training, governance, and multidisciplinary collaboration. In Spanish: https://ucontinental.edu.pe/investiga/manifiestoia/ The added value of the AI Dialogue lies in its ability to connect these fragmented efforts into a coherent global ecosystem—linking policy, practice, and capacity-building. By doing so, it can accelerate the transition from isolated initiatives to scalable, interoperable, and ethically grounded AI systems, ensuring that innovation advances as a global public good.

An effective AI Dialogue must be designed as a multi-level, inclusive, and action-oriented process, where diverse stakeholders contribute not only perspectives, but also solutions, commitments, and implementation pathways.

Global discussions on AI governance continue to underrepresent several critical voices—particularly those most affected by AI deployment but least involved in its design and regulation. Low- and middle-income countries (LMICs) remain underrepresented, especially from regions such as Latin America, Africa, and parts of Southeast Asia. Their participation is often limited to consultation rather than co-decision-making, despite facing the greatest implementation challenges and risks of dependency on external technologies.

To move beyond traditional, passive consultations, the AI Dialogue should adopt innovative, participatory formats that prioritize co-creation, real-world problem-solving, and sustained interaction. Regional "deep dive" dialogues are essential to ensure contextual relevance. These smaller, facilitated sessions—particularly in LMIC settings—can surface local priorities, cultural dimensions, and system constraints that are often overlooked in global forums. Fourth, multi-stakeholder challenge sprints or hackathons can mobilize interdisciplinary teams to address specific governance questions (e.g., bias detection, explainability tools, data-sharing frameworks), generating rapid, practical solutions.

Capacity-building initiatives such as regional digital health and AI training hubs are essential. Evidence shows that embedding AI competencies into medical, public health, and engineering curricula is key to sustainable adoption and governance. Importantly, context-specific approaches are emerging. For example, the Arequipa Manifesto 2025 on the Ethical Use of AI in Medical Education promotes a person-centered, ethics-driven, and competency-based framework, emphasizing that AI should augment-not replace-clinical judgment, while ensuring data protection, equity, and continuous professional development (https://ucontinental.edu.pe/investiga/manifiestoia/). Together, these examples highlight that effective AI governance is not a single policy, but an ecosystem of aligned frameworks, experimentation mechanisms, open collaboration, and sustained capacity-building, adapted to local contexts while connected globally.