UN leaders are concerned about AI’s potential harms. Could an AI forum prevent the worst?

Artificial intelligence (AI) took  center stage at this week’s annual high-level United Nations (UN) meeting in New York.

Leaders at the UN Security Council addressed AI’s possible benefits and harms in security, military use and misinformation.

“The question is not whether AI will influence international peace and security, but how we will shape its influence used responsibly,” UN Secretary-General Antonio Guterres said in opening remarks at Wednesday’s meeting.

“AI can strengthen prevention and protection, anticipating food insecurity and displacement, supporting de-mining, helping identify potential outbreaks of violence, and so much more. But without guardrails, it can also be weaponised,” Guterres added.

Wednesday's general debate centred around how the Council can help ensure the responsible application of AI to comply with international law and support peace processes and conflict prevention.

How have world leaders reacted?

Several European leaders stressed the need for the Council to lead the way on ensuring that AI is not used by militaries without human oversight to avert potentially devastating escalations or misfires.

Greek Prime Minister Kyriakos Mitsotakis called on the Council to “rise to the occasion – just as it once rose to meet the challenges of nuclear weapons or peacekeeping, so too now it must rise to govern the age of AI.”

British Deputy Prime Minister David Lammy stressed deep AI analysis of situation data  holds a promise for peace, saying AI is capable of keeping “ultra-accurate, real-time logistics, ultra-accurate real-time sentiment analysis, ultra-early warning systems”.

UN sets up new bodies for AI

Last month, the UN General Assembly (UNGA) announced that it will set up two key bodies on AI - an independent scientific panel of experts and a global forum.

The UN said in a statement that the new governance architecture will be a much more inclusive form of international governance and address the issues surrounding AI, and ensure that it benefits all people.

The Scientific Panel, for which 40 experts will be appointed through nominations, will present annual reports at the forum named Global Dialogue on AI Governance to take place in 2026 in Geneva and 2027 in New York.

The new establishment is seen as the latest and biggest effort to rein in AI. Experts have called it “a symbolic triumph”.

They are “by far the world’s most globally inclusive approach to governing AI,” Isabella Wilkinson, a research fellow at the London-based think tank Chatham House, wrote in a blog post.

Britain, France, and South Korea have all held global AI summits but none of them have resulted in binding pledges for AI safety.

However, Wilkinson is sceptical that the UN's lumbering administration can regulate a fast-moving technology such as AI.

“But in practice, the new mechanisms look like they will be mostly powerless,” she added.

The UN chief will hold a meeting to officially launch the two new bodies on Thursday. It will be the first time that all 193 Member States of the UN will have a say in the way international AI governance is developed, according to the UN.

Previously, leading AI experts and Nobel Prize winners, including senior figures from OpenAI, Google DeepMind and Anthropic, had issued a call for the United Nations to spearhead a binding global treaty setting “minimum guardrails” for AI designed to prevent the “most urgent and unacceptable risks”.

Among those who signed the call were European lawmakers, including Italian former prime minister Enrico Letta and former president of Ireland Mary Robinson, who is currently a United Nations high commissioner for human rights.