HealthAI Contributes to Shaping the UN Global Dialogue on AI Governance
Introduction
HealthAI - The Global Agency for Responsible AI in Health, recently submitted recommendations to the formal consultation on the UN Global Dialogue on AI Governance, advocating for a coordinated, multistakeholder, and mission-driven approach to global AI governance.
Background on the UN Global Dialogue Transformation
The United Nations’ Digital Compact has been working on the crucial elements that will shape two new mechanisms to advance global AI governance: An international scientific panel on AI and the UN Global Dialogue on AI Governance. The initiatives were proposed by the High-Level Advisory Body on AI in their Governing AI for Humanity report, released in September 2024.
The report, authored by world-renowned experts, recommends establishing a twice-yearly intergovernmental and multi-stakeholder policy dialogue on AI governance. This forum aims to share best practices, promote common understanding, and address AI incidents, facilitating open, transparent, and inclusive discussions on AI governance at a global level.
Following the launch of the report and approval under the “Pact for the Future” (UN Resolution 79/1 of 22 September 2024), the Secretary-General of the UN launched an intergovernmental process and formal multi-stakeholder consultations to delineate the scope and terms of the two mechanisms mentioned above. HealthAI submitted recommendations on the establishment of the UN Global Dialogue on AI Governance, specifically.
HealthAI's Vision for the Global Dialogue on AI Governance
In our submission, we outlined a strategic vision for the scope and authority of the Global Dialogue. We believe the Dialogue should serve as a central institutional coordination convening for all AI governance matters in the UN system. Its authority should be derived from robust multi-stakeholder participation. Based on multi-stakeholder deliberation, the Dialogue should be responsible for defining AI governance priorities, setting the agenda, and coordinating work within the UN system. Finally, regarding substance, we recommended that the Dialogue's AI governance mandate be closely aligned with the Sustainable Development Goals' agenda and timelines.
Three Priority Outcomes
HealthAI identified three priority outcomes that the Global Dialogue should strive to achieve:
1. Unify UN Efforts for Global AI Governance
The Dialogue should work toward unifying UN efforts to host a flagship global AI governance forum. This can be accomplished through the coordination among existing convenings at the UN system focused on AI, such as ITU’s AI for Good Global Summit and UNESCO’s Global Forum on the Ethics of AI. This unified flagship initiative would set the agenda for a series of concrete commitments, such as the collective development of blueprints—including guidelines, resolutions, reports, and frameworks. We expect that agreements stemming from the Dialogue will be based on a new coordination paradigm, which is laid out in the two points below.
2. End the AI Race to the Bottom
The Dialogue presents a unique opportunity to shift the competitive geopolitical paradigm toward productive discussions focused on solving collective problems caused or exacerbated by the scaling of AI. Aiming further than existing commendable "like-minded" approaches to international AI governance (such as the OECD.AI, the G7 Hiroshima AI Process, and the Council of Europe Framework Convention on AI and Human Rights), the Dialogue should focus on building bridges between actors currently engaged in competitive races for AI dominance. In addition, it should include underrepresented regions in the AI governance agenda.
Given the UN's broad membership and track record in complex international coordination issues like climate change, the Global AI Dialogue is arguably the only forum that can shift AI governance toward a truly collaborative approach. Notably, UNESCO's Recommendation on the Ethics of AI, adopted in 2021 and applicable to 194 countries, serves as evidence of the UN's unique ability to achieve broad consensus on AI matters. We believe that the UN can replicate the Recommendation’s success in the Global Dialogue initiative.
3. Implement a Problem-Solving Approach
The Dialogue should unite the international community to tackle concrete issues, particularly AI's impacts in specific sectors. By focusing on specific issues at a time—such as AI in health—the Dialogue can foster pragmatism and reconcile political divides. Additionally, bringing civil society to the center stage with a novel structure for international coordination may help diffuse tensions among state actors with competing national agendas and steer discussions toward socially beneficial outcomes.
Civil Society Participation
HealthAI emphasized the critical importance of active participation by civil society organizations in the Global Dialogue. The Dialogue should take the opportunity to advance the UN's tradition of multi-stakeholder engagement by granting civil society representatives a central role and fostering cultural and geographic diversity in civil society participation. We specifically highlighted that the UN should boost the representation of civil society organizations from the Global South.
A Mission-Driven Approach
Finally, HealthAI proposed a specific format for the Dialogue. We urged the Dialogue to reclaim a central role in international coordination by promoting a cross-sectoral, long-term vision for global AI governance that balances risks and opportunities. To achieve this vision, the Dialogue should adopt a format akin to a "mission-driven" approach, where stakeholders discuss forward-looking strategies to tackle concrete issues related to the intersection of AI and SDG themes such as health, infrastructure, and education. For example, health-related discussions might strategize the establishment of a global incident reporting mechanism for AI deployment in the health sector. They could also focus on leveraging AI to expand access to primary healthcare and improve the quality of healthcare in regions where doctor-to-patient ratios are historically high.
This mission-driven approach offers a proactive framing, encouraging discussions to focus on collectively shaping how AI should be deployed and what policies are needed to steer it toward socially beneficial outcomes. Given the rapid pace of AI development and adoption, it is urgent to move beyond principle-level discussions and explore formats that encourage hands-on collaboration rather than competition.
HealthAI remains committed to advancing responsible AI governance in healthcare and will continue to engage with international initiatives like the UN Global Dialogue on AI Governance. Follow our work and updates on further developments in this space.
