The United Nations is entering its annual high-level week with an unusually crowded test of its relevance. Nearly 130 national leaders are expected in New York as wars, economic tension, climate risks and artificial intelligence force governments to ask how much can still be managed through institutions built for a different era.
The 81st session of the General Assembly formally opened on September 9. Its high-level week runs from September 18 to 28, with the general debate scheduled for September 22-26 and September 28. The session is being led by General Assembly President Khalilur Rahman of Bangladesh under the theme “Restoring trust, managing transformation: A United Nations that delivers for all.”
The theme is deliberately institutional. The United Nations is not confronting only a list of crises. It is confronting doubt about whether multilateral institutions can still produce collective action when major powers disagree.
War remains the immediate test
The General Assembly does not command armies and most of its resolutions are not legally binding. Yet it remains the only global political body in which all 193 member states sit with one vote each. That gives it a different kind of authority: states that cannot agree in the Security Council can still debate, organize coalitions and establish broad political positions in the Assembly.
That role becomes more visible when the Security Council is divided. The wars involving Ukraine, the Middle East and other regions have repeatedly exposed the limits of consensus among permanent members. General Assembly votes cannot by themselves end those conflicts, but they document where international support lies and provide political frameworks for humanitarian, legal and diplomatic action.
This year’s high-level week is therefore likely to contain the familiar contrast between ambitious speeches and limited enforcement capacity. Leaders will use the podium to define threats, assign responsibility and seek allies. The practical question is what happens after the speeches.
A United Nations under financial pressure
The institution also faces a material problem. U.S. funding cuts and broader budget pressures have intensified questions about what the UN can realistically deliver. Reuters reported that President Donald Trump, who is scheduled to address the Assembly, has reduced U.S. support to parts of the UN system and criticized the institution’s effectiveness.
Because the United States is one of the largest contributors to the UN system, changes in Washington’s funding policy have consequences beyond diplomatic symbolism. Peacekeeping, humanitarian programmes and specialized agencies depend on assessed and voluntary contributions from member states.
The debate is therefore not simply whether multilateralism is desirable. It is whether governments are willing to finance the institutions they say they need.
Artificial intelligence moves into security diplomacy
The newest major issue is artificial intelligence. AI governance was once discussed mainly by technology ministries, standards bodies and specialist forums. It is now moving into the core institutions of international security.
OpenAI chief executive Sam Altman is scheduled to brief the UN Security Council during high-level week in a meeting convened by France on the international-security implications of AI. Reuters reported that the discussion is expected to cover shared safety standards, dangerous uses of advanced systems and the need for international coordination.
The General Assembly presidency has already made AI a priority. At an ambassadorial lunch on September 10, Rahman argued that the UN should help ensure that AI serves human development rather than becoming a source of instability or harm. He pointed to continuing work on the Global Digital Compact, the Global Dialogue on AI Governance and scientific advisory mechanisms.
The logic is simple: advanced AI systems are being developed by a small number of companies and states, but their effects can cross borders immediately. Cyber operations, biological design, military decision support, financial markets and information systems create risks that do not fit neatly inside national regulation.
The representation problem
AI also creates a political problem for global governance. The countries leading frontier-model development are not representative of the world as a whole. Many developing states worry that rules could be written by a small group of technological powers and then exported globally.
The General Assembly’s advantage is universality. Every member state has a formal voice. Its disadvantage is that universal participation can make agreement slow and general. The Security Council has more concentrated political weight but reflects the power structure of 1945, not the distribution of technological capability in 2026.
That tension explains why the UN is building several overlapping AI processes rather than one global regulator. Scientific panels can establish shared evidence. Dialogues can build norms. The General Assembly can produce political commitments. National governments will still implement most binding rules.
Climate, development and technology converge
High-level week will not be only about war and AI. The official UN programme includes climate change, pandemic preparedness, sea-level rise, racism and progress toward the Sustainable Development Goals. These issues increasingly overlap with security and economic policy.
Climate shocks affect migration, food prices and state budgets. Pandemic preparedness requires supply chains and data sharing. AI can improve health, agriculture and public administration while also increasing inequality or concentrating power. Development finance is linked to debt and interest rates. The agenda is crowded because the problems themselves are connected.
The problem of trust
The Assembly president’s emphasis on restoring trust is therefore more than rhetoric. Multilateral institutions work only when states believe rules will be applied with enough consistency to justify cooperation. Selective enforcement, unresolved wars and repeated Security Council vetoes weaken that belief.
But distrust does not automatically make international institutions irrelevant. In some cases it makes them more necessary. Rival states still need places to exchange messages, negotiate technical standards and build agreements on issues where interests overlap.
The U.S.-China AI discussions taking place elsewhere in New York this week illustrate the same point. Competition between major powers does not eliminate the need for guardrails. It increases it.
What high-level week can realistically do
The General Assembly is unlikely to resolve the world’s largest conflicts in one week. It can, however, clarify positions, launch negotiations, mobilize funding, establish new processes and create political pressure that continues after leaders leave New York.
The strongest test will be follow-through. An AI declaration matters if it creates mechanisms that governments and companies actually use. A development pledge matters if financing follows. A peace initiative matters if parties return to negotiations.
The UN was designed after a catastrophic world war around the idea that even adversarial states need common institutions. In 2026, the question is not whether the world resembles 1945. It plainly does not. The question is whether the system can adapt quickly enough to wars, technologies and economic pressures its founders could not have anticipated.
High-level week begins with that question unresolved. The institution’s value will be measured less by the number of speeches delivered than by whether governments can still turn shared problems into shared rules.