User Experience Philippines and SIGCHI Manila

This meeting should NOT be mere symbolic. It should have some "teeth" to it precise action items and timelines moving forward. First, meaningful inclusion of the Global South. Addressing AI capacity gaps in developing countries is one of the Dialogue's stated aims, and for good reason — AI is rapidly concentrating economic power among a handful of wealthy nations and corporations. Success means Global South countries leaving Geneva not just as participants, but as co-architects of norms that actually reflect their development needs and risk profiles. Second, a credible scientific baseline. Without a common baseline, fragmentation wins — with different regions operating under incompatible policies and technical standards. The Independent Scientific Panel was established precisely to provide shared facts. For the Dialogue to succeed, governments need to actually anchor their positions in that evidence, rather than treating it as optional decoration. Third, convergence on transparency and accountability principles, even where binding rules remain out of reach. Meaningful human oversight in every high-stakes decision — in justice, healthcare, credit — and clear accountability so responsibility is never outsourced to an algorithm should be minimum agreed ground. Finally, a credible roadmap forward. The first Dialogue sets the tone for all that follow. Its success depends on producing not just a communiqué, but a sequenced, actionable agenda — one that makes it harder, not easier, for major powers to opt out over time. The honest risk is that it becomes, as the Atlantic Council warns, "global in form but geopolitical in substance." Avoiding that outcome is itself a success condition.

• Safe, secure and trustworthy AI
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights
• Transparency, accountability, and human oversight

The first Global Dialogue on AI Governance is a rare opportunity to establish foundational principles before harmful precedents become entrenched. Safety and security must anchor all frameworks. As AI is deployed in healthcare, critical infrastructure, and national security, shared safety standards prevent a race to the bottom among competing jurisdictions. Inclusion and cultural perspectives are core legitimacy requirements, not optional additions. Governance frameworks that exclude the values and lived realities of the Global South, Indigenous communities, and marginalized populations will entrench existing inequalities. Genuine inclusion means influence over outcomes, not merely a seat at the table. Ethics grounded in human rights provides the universal floor beneath which no jurisdiction should fall. Anchoring AI governance in internationally recognized human rights frameworks ensures continuity with existing international law and prevents capture by purely commercial or geopolitical interests. Accountability with human oversight closes the gap between principles and practice. High-stakes decisions affecting people's lives must remain subject to human judgment and meaningful redress - not delegated irreversibly to automated systems. These priorities form the foundation for AI governance that is trustworthy, equitable, and durable.

Environmental sustainability is conspicuously absent. The resource intensity of large AI systems - energy consumption, water usage, and e-waste from hardware cycles - has significant implications for climate commitments and environmental justice, particularly in regions already bearing disproportionate climate burdens. AI governance frameworks must account for ecological cost alongside human impact. AI and conflict deserves explicit treatment beyond security framing. The use of AI in warfare, autonomous weapons systems, and military targeting raises profound humanitarian law questions that existing themes touch only obliquely. Without dedicated attention, governance frameworks risk legitimizing uses of AI that violate the laws of armed conflict. Epistemic rights and information integrity cut across transparency, human rights, and ethics but are fully captured by none. AI-generated disinformation, synthetic media, and algorithmic curation of information environments threaten people's ability to form beliefs freely - a foundational condition for democratic participation and informed consent. Intergenerational and future-oriented perspectives are underrepresented. Current governance discussions are dominated by present-state harms and near-term risks. Frameworks that do not explicitly consider long-term, potentially irreversible consequences - particularly for children and future generations - will be structurally incomplete. These issues are not peripheral. They are the terrain where today's governance choices will have the most lasting consequences, and they deserve explicit recognition in the Dialogue's agenda rather than treatment as afterthoughts within existing themes.

Impact on the Medical Device Sector — United States and Philippines AI governance gaps are having concrete and consequential effects on the medical device sector in both the United States and the Philippines, with distinct but interconnected challenges. In the United States, the FDA has made meaningful progress through its AI-enabled device framework and predetermined change control plans, but significant gaps remain. Transparency and accountability standards for AI-driven clinical decision support tools are inconsistent, leaving patients and clinicians uncertain about how recommendations are generated or how errors are caught. Human oversight requirements are unevenly applied, particularly for adaptive AI systems that change behavior post-market. The absence of harmonized international standards also creates regulatory fragmentation that burdens manufacturers and slows safe innovation. In the Philippines, the challenges are more foundational. Limited regulatory infrastructure, AI capacity, and technical expertise mean that AI-powered medical devices — many imported — are evaluated without robust frameworks for algorithmic accountability, bias assessment, or post-market surveillance. Devices trained predominantly on Western patient populations may perform poorly on Filipino patients, embedding health disparities directly into clinical tools. This is a human rights concern as much as a technical one. Across both contexts, the lack of interoperable governance approaches creates a two-tiered system where high-income markets drive AI development standards while lower-resource settings absorb the risks. For medical devices specifically, this manifests in diagnostic tools, imaging AI, and clinical decision support systems that may be deployed without adequate validation for local populations, disease profiles, or clinical workflows. The opportunity lies in establishing sector-specific AI governance guidance — covering safety, transparency, human oversight, and equity — that is both globally interoperable and locally adaptable, ensuring that advances in medical AI benefit all patients, not only those in well-resourced health systems.

The Role of the AI Dialogue in Advancing International Cooperation As someone working at the intersection of medical device development, usability engineering, and regulatory compliance across the United States and the Philippines, I see the Dialogue's potential through a concrete lens. Harmonizing fragmented regulatory frameworks is the most immediate opportunity. Navigating FDA AI guidance alongside EU MDR already creates significant compliance complexity for medical device manufacturers. When Philippine regulatory capacity enters the equation, smaller markets default to accepting standards they had no role in shaping. The Dialogue can drive convergence on core principles — transparency, human oversight, post-market accountability — without demanding identical rules across jurisdictions. Centering equity in health AI is equally urgent. AI-powered diagnostic and clinical decision support tools are predominantly trained on Western patient populations. Filipino patients and health systems deserve governance frameworks that require validation across diverse populations before deployment, not after harm occurs. Building genuine regulatory capacity in lower-resource settings must move beyond rhetoric. The Philippines needs not just access to governance frameworks but meaningful support developing the technical expertise to evaluate AI-enabled medical devices rigorously. International cooperation here directly protects patients. Finally, the Dialogue must produce accountability mechanisms, not just declarations. In a field governed by standards like IEC 62366-1 and FDA human factors guidance, practitioners understand that principles without verification processes are insufficient. The same discipline must apply to international AI governance commitments — structured review, measurable benchmarks, and consequences for inaction. The patients at the end of these systems deserve nothing less.

Several existing frameworks are directly relevant to the medical device sector and to the regulatory realities of the United States and the Philippines. The FDA's AI-Enabled Medical Device framework and its predetermined change control guidance represent some of the most sector-specific AI governance work globally. The Dialogue should connect with this work to extract applicable principles — particularly around post-market surveillance and adaptive algorithm oversight — and internationalize them for broader adoption. IEC 62366-1 and the broader IEC/ISO medical device standards ecosystem already provide internationally recognized frameworks for usability engineering and human factors. These standards embed human oversight principles that AI governance frameworks frequently reinvent from scratch. The Dialogue should explicitly build on this existing infrastructure rather than creating parallel obligations. The WHO's guidance on ethics and governance of AI for health provides a human rights-grounded, globally inclusive foundation specifically for health AI. This is particularly relevant for the Philippines and other lower-resource settings where health AI deployment is outpacing regulatory readiness. The ASEAN Guide on AI Governance and Ethics reflects regional efforts relevant to the Philippine context. The Dialogue can add value by connecting ASEAN-level commitments to global frameworks, ensuring regional priorities — including linguistic diversity, infrastructure constraints, and cross-border data flows — are not lost in translation. The added value the Dialogue brings is integration. Each of these initiatives operates within its own silo — sectoral, regional, or thematic. The Dialogue's unique contribution is serving as the venue where health sector practitioners, human rights advocates, regulatory authorities, and lower-resource country representatives converge around shared, actionable commitments — with accountability mechanisms that existing initiatives largely lack.

Panel Speaker opportunity both short and long-form TED-style talks Working groups composed of global participants that can meet regularly in person or online

Global South Minority groups in the US (women, people of color) Design-oriented disciplines need to be included Psychologists like myself - who understands human behavior

Online and in-person meetings (hybrid and more frequent)

The medical device sector offers some of the most mature and instructive models for AI governance - built through decades of hard-won regulatory experience balancing innovation with patient safety. The FDA's Predetermined Change Control Plan (PCCP) is a concrete governance innovation worth elevating globally. It requires manufacturers to specify in advance how AI algorithms may change post-market, under what conditions, and with what validation - embedding transparency and human oversight directly into the product lifecycle rather than treating them as afterthoughts. IEC 62366-1 and formative/summative usability testing frameworks demonstrate that human-centered design principles can be operationalized rigorously and verifiably. Requiring evidence that real users - including clinicians with varying expertise levels - can safely and effectively interact with AI-driven interfaces is a transferable governance model applicable well beyond medical devices. The EU MDR's post-market clinical follow-up requirements establish ongoing accountability obligations that prevent governance from ending at market approval. This continuous surveillance model - requiring manufacturers to monitor real-world performance and act on emerging safety signals - is a practical accountability mechanism that AI governance frameworks across sectors should adopt. The WHO's Ethics and Governance of AI for Health guidance demonstrates that principles can be made actionable when developed with genuine multi-country participation, including lower-resource settings. Its emphasis on inclusivity, transparency, and accountability provides a replicable process model. At the platform level, structured adverse event reporting systems - analogous to the FDA's MAUDE database - offer a proven approach for systematically capturing AI-related failures in deployment, enabling evidence-based governance iteration over time. The common thread across these examples is that effective governance is specific, verifiable, and continuous - not aspirational and static.