Global Gaps in AI Governance

According to the UN High-Level Advisory Body on AI, seven countries currently participate in all major global AI governance initiatives. By contrast, 118 countries do not participate in any. Various UN agencies, such as UNESCO, the International Telecommunication Union (ITU), and the UN Development Programme (UNDP) have developed their own AI principles and capacity-building efforts. To date, these initiatives have not evolved into a unified global governance system.

The Formulation of a Strategy

The process of building a globally coherent and inclusive AI governance structure required multiple consensus-building negotiations. The Summit of the Future, held in September 2024, called for strengthening international cooperation on AI and creating a strong AI governance structure. Feeding into the negotiation, the final report Governing AI for Humanity was published by the UN in December 2024. The report proposed the creation of both a scientific panel and a global policy dialogue. Together, the two mechanisms were designed to combine evidence-based technical expertise with inclusive political engagement.

Reaching Agreement in 2025

Over the following months, member states discussed how to turn the recommendations of the report into a structured system. This process reached agreement by consensus in August 2025 on the AI Modalities Resolution (A/RES/79/325). This resolution formally established the two mechanisms: Independent International Scientific Panel on AI and the Global Dialogue on AI Governance.

The AI Modalities Resolution

Ultimately, on 26 August 2025, UN member states adopted the AI Modalities Resolution (A/RES/79/325) establishing the Independent International Scientific Panel on AI and the Global Dialogue on AI Governance. This created a space where experts and countries could engage collectively on emerging AI governance principles.

Mandate and Composition of the Scientific Panel

According to the AI Modalities Resolution, the Panel is intended to “provide authoritative, evidence-based scientific assessments”. These are meant to help member states understand AI developments and implications. The resolution mandates a panel of 40 independent experts, selected with geographical and gender balance to ensure diverse representation. The Panel members are nominated through a process coordinated by the UN Secretary-General. The appointments are based on scientific and technical expertise, and chosen members serve fixed terms. This structure aims to maintain continuity while allowing periodic renewal of expertise.

Scientific Advice for Policymaking

The Panel produces annual evidence-based assessments of AI systems, their capabilities, and potential impacts to inform policymaking. Under the resolution, members serve in their personal capacity and are subject to conflict-of-interest rules. As described in the resolution, the Panel provides analyses from their research and technical expertise to the Global Dialogue. This system creates cooperative relationships between the two mechanisms, allowing scientific research to inform policy discussions.

A Global Forum on AI

The Global Dialogue provides a forum for states, experts, and stakeholders to discuss AI trends, policy responses, and identify issues that require global cooperation. According to the Tech Policy Press, the Global Dialogue is the first global, inclusive platform with broad participation including countries from the Global South and LDCs. It holds annual meetings, beginning with the launch at UNGA 80 (September 2025), with the next one scheduled to take place in Geneva in 2026. The Global Dialogue’s agenda is created based on discussion by the member states and stakeholders on emerging AI governance issues requiring international consideration.

Including Global South Voices

The results of the Dialogue are organized into a summary report outlining key themes and arguments, without legally binding decisions. According to an analysis published by Tech Policy Press, the mechanism is described as a platform to bring North–South perspectives together. This analysis addresses an issue that many UN processes have historically struggled to include: The perspectives of the Global South. The mechanism is structured to provide equal opportunities for all countries to participate in shaping international AI governance. It complements existing forums by providing a space for countries to engage in discussions. Within this framework, initiatives developed in bodies such as the IGF, G7, OECD, AU and the Council of Europe can be compared and discussed.

Financing the Mechanisms

Now that the mechanisms have been established, its effectiveness will depend on how operational challenges are addressed. One key challenge concerns financing, which has been an unresolved issue in public-interest cooperation according to a Chatham House analysis. The UN’s tight budget constraints, due to competing demands across the United Nations and reduced funding from the United States, could complicate efforts to secure sustainable funding for the work of AI governance. According to the Secretary-General’s report (A/79/966), large-scale investment will be needed to support the panel and the Global Dialogue. There is also a risk of influence from the funders who may seek to reflect their own interests.

Limits of Voluntary Governance

A legal briefing by the law firm White & Case also notes that the mechanisms rely on voluntary participation rather than enforcement mechanisms. The panel’s role is strictly advisory. It is not meant to set regulatory standards or implement policy recommendations. The decision-making is solely up to the member states and other institutional bodies. While this reflects a common limitation across many UN conventions, it also reflects that sovereignty lies with member states and their people.

Towards Coordinated AI Governance

As AI technologies expand across sectors, they bring both practical benefits and risks, including uneven access and the possibility of technology outpacing governance. In response, the United Nations has strengthened its role by creating two mechanisms. Together, these mechanisms reflect a shift towards a more inclusive and structured approach to global AI governance, grounded in scientific assessment. With the previous efforts by various agencies, organizations, and governments, the recent development fits into the global governance picture. The next stages will test inclusivity, credibility, and cooperation.