The UN Advances Global AI Governance

The UN AI Modalities Resolution, adopted by consensus, marks a significant step toward establishing a framework for global AI governance. The resolution, A/RES/79/325, sets out the terms for two new institutional mechanisms aimed at steering AI development and deployment in a manner that is safe, ethical, and beneficial to all. However, as noted by Vidisha Mishra and Nicole Manger of the Global Solutions Initiative, the success of these mechanisms hinges on addressing several practical challenges, primarily related to independence, resources, and inclusive participation.

Two New Institutional Mechanisms

Independent International Scientific Panel on AI: This panel will be composed of 40 experts, chosen for their personal capacity with a balanced representation of gender and geography. Its primary task is to produce an annual, evidence-based assessment that synthesizes existing research on the opportunities, risks, and impacts of AI. This report is designed to be “policy-relevant but non-prescriptive,” providing a scientific foundation for political discussions on AI governance. A key feature is the requirement for members to disclose conflicts of interest, a measure intended to protect the panel’s credibility.

Global Dialogue on AI Governance: This is a truly global, multi-stakeholder platform that will convene annually within the UN framework. Its purpose is to facilitate open and inclusive discussions among governments, the private sector, civil society, and other relevant stakeholders. The dialogue aims to promote international cooperation, share best practices, and address key issues such as closing the digital divide, building AI capacity in developing countries, and ensuring the human rights, ethical, and social implications of AI are addressed.

The Three Critical Challenges

According to Mishra and Manger, the effectiveness of these new bodies is not guaranteed and will depend on overcoming three critical hurdles:

Independence: The authors argue that the credibility of both the Scientific Panel and the Global Dialogue rests on their ability to operate independently, free from undue political or corporate influence. The resolution’s call for full disclosure of conflicts of interest is an important safeguard. However, the authors emphasize that this must be robustly enforced to prevent powerful interests from exerting quiet control over the selection of experts or the direction of the dialogue.

Resources: The resolution’s heavy reliance on voluntary funding poses a significant challenge. With the UN’s regular budget under constraint, the new mechanisms could be financially vulnerable. Mishra and Manger stress that predictable and sustainable funding is essential, not only to ensure the operational viability of the new bodies but also to prevent financial power from translating into political influence. They suggest that new funding models, such as a Global Fund on AI, could be explored to prevent conflicts of interest and ensure that all countries, particularly developing ones, have the resources to participate meaningfully.

Inclusive Participation: The success of the resolution is tied to its promise of inclusivity, especially for countries and communities from the Global South that have been historically excluded from AI governance discussions. The authors warn that without a clear and enforceable framework for representation, the new mechanisms risk becoming dominated by a narrow set of perspectives. They highlight the need for deliberate efforts to ensure that women, people with disabilities, youth, and other marginalized groups are not just present, but are empowered to shape the agenda and outcomes of the dialogue. The resolution encourages travel support for participants from developing countries, a critical step to bridge the gap in access.