By Javier Surasky
Keeping track of everything happening in AI is almost impossible. New developments emerge nearly every day, driven by debates on technological, ethical, and legal issues, among others.
This blog post, therefore, seeks to bring some order to what has happened since the start of the year solely within the United Nations system, which—weak as it is—remains the best forum we have for advancing shared governance of artificial intelligence.
The decision to publish this kind of chronology is tied to the upcoming start of the General Debate, where we expect AI to feature in several statements—an issue we have addressed previously on this blog.
We have selected a set of moments that seem to us the most relevant within a list that could certainly be longer, but which offers a good, holistic, and comprehensive view of the discussions and the contributions from the global framework.
Here is our selection, in chronological order:
January 1: The ODET (Office for Digital and Emerging Technologies) is established.
The Office of the Secretary-General's Envoy on Technology is transformed into ODET, mandated to coordinate implementation of the Global Digital Compact and, within it, AI governance processes, becoming the Secretariat's focal point on the issue.
February 1: Canada proposes delimiting discussions on AI and cybersecurity within the UN process.
February 3: The High-Level Conference on AI and Work is held by the ILO and the European Economic and Social Committee (EESC), focused on the principles of social justice, social dialogue and safeguards regarding AI deployment in employment, as part of an ongoing process to build a common language between the ILO and the European Union on AI's labor risks and opportunities.
February 10–11: AI Action Summit in Paris, launching the Coalition for Sustainable AI led by France, UNEP, and ITU. Positions among the United States, China, and the EU on AI regulation are clearly on display. A Plan of Action for Global Governance of Artificial Intelligence (A/79/980) is adopted there.
April 4: A Special Meeting of the Security Council on AI for peace and security takes place.
April 7: UNCTAD releases its flagship report, "Technology and Innovation Report. Inclusive Artificial Intelligence for Development." The report calls for global cooperation to make AI inclusive, warns of geographic/corporate concentration, and proposes policy levers such as infrastructure development, data, talent, and standards to promote equity, while seeking to frame the debate on capacity-building and development financing for AI.
May 6: ECOSOC holds a special meeting on AI.
May 20: The ILO launches a refined global index of exposure to AI (Working Paper No. 140, WP-140), updating its measurement of occupational exposure to Generative AI. It concludes that AI transforms more jobs than it automates—central evidence for designing training, social protection, and maintaining collective labor bargaining.
June 25: The 68th session of UN-COPUOS adopts international recommendations on AI regulation in outer space, based on safety, responsibility, and analogies with similar regimes. This brings to COPUOS, for the first time, a thematic package devoted exclusively to AI.
July 1: The Human Rights Council adopts its resolution on human rights and new digital technologies (A/HRC/59/L.14)
July 2: Responding to a request from the countries that participated in the Fourth International Conference on Financing for Development, the Secretary-General issues his report, Innovative voluntary financing options to build capacity in artificial intelligence (A/79/966)
On the same day, UNCTAD publishes the report "Implementing World Summit on the Information Society Outcomes: a 20-year review", systematizing advances and gaps in the information society, with emphasis on AI, open data, and infrastructure.
July 7–11, 2025: AI for Good 2025 (ITU) convenes, focusing on digital innovation, governance, and the establishment of shared standards.
July 7: The United Nations International Computing Centre (UNICC) inaugurates its "AI Hub" during WSIS+20/AI for Good as an inter-agency platform for governance, collaboration, and working tracks on common UN AI needs, to share solutions and reduce duplication.
July 28: The United Nations University launches the "Global AI Network" and announces the AI for SDGs — Global Youth AI Future Innovation Competition 2025, focused on SDG-oriented applications. (See: unu.edu)
August 4: UNCTAD publishes Policy Brief No. 120, "Preparing to seize artificial intelligence opportunities with strategic national policies", presenting tools for developing national AI strategies (sectoral priorities, data, norms, public procurement), helping countries that seek to translate AI capabilities and governance into concrete roadmaps.
August 14: The ILO launches its "Global policy tracker on digital labour platforms" to "support evidence-based policymaking and social dialogue on platform work."
August 26: The General Assembly adopts resolution 79/325 establishing an International Scientific Panel on Artificial Intelligence and launching the annual Global Dialogue on AI. This is the most significant advance within the period analyzed.
In addition, as the culmination of a process begun in 2023, the United Nations University sets up the UNU-AI Centre in Bologna, focused on big data and AI, which will work closely with UNU's headquarters in Macau (China), where a hub on the topic already exists and which—commemorating the UN's 80th anniversary and the 50th anniversary of UNU—will host the "AI for Humanity" Conference from October 23 to 25 of this year. These advances strengthen academic capacities in AI within the UN system.
What lies ahead for the rest of this year?
On September 24, the Security Council will hold a high-level open debate on "Artificial Intelligence and the maintenance of international peace and security."
On December 16–17, the General Assembly will host the WSIS+20 High-Level Meeting for the final review of the World Summit on the Information Society. Its agenda includes several chapters related to AI (digital governance and capacity-building in AI, among others).
No date has been announced yet, but before the end of the year, a public call for nominations to serve on the International Scientific Panel on AI is expected, and the process will begin to launch the first Global Dialogue on AI Governance.
It is also expected that the specific proposals the Secretary-General must submit to the General Assembly as part of the "reform" initiative known as UN80 will include several issues related to implementing the Global Digital Compact and strengthening the use of AI in the UN's institutional management.
The main overall conclusions that emerge from this overview are:
- The UN is expanding its institutional architecture to address the risks and opportunities arising from the massification and advancement of AI. Still, a risk of duplication is beginning to surface—a recurring problem in the Organization's structure.
- At the same time, incipient yet determined steps are being taken to build an academic underpinning for the study of AI, including both the creation of the expert Panel and the developments within the United Nations University. It remains to be seen how these spaces will be articulated among themselves.
- The model being created appears to replicate—or at least serve as a basis for—the Organization's working model on climate change. The expert Panel could end up becoming a kind of IPCC, with the attendant risk of turning into an evidence provider without the capacity for the transformative action required.
- There is an observable assumption that AI could solve several management problems within the UN and support countries' growth, but neither will happen naturally. Sustained political support is required for both opportunities to become realities.