Singapore Calls for Global AI Safeguards at the UN General Assembly

On 26 September 2026 Singapore’s Foreign Minister Vivian Balakrishnan delivered the nation’s statement at the 81st United Nations General Assembly in New York. In a 20-minute speech he warned that artificial-intelligence (AI) capabilities are outpacing the world’s ability to manage their risks and urged the creation of new international rules and institutions, including a possible UN Framework Convention on AI Safeguards.

Singapore proposes a UN framework convention on AI safety At UN, Singapore warns AI safeguards must keep pace as capabilities race ahead The Jerusalem PostChannel News Asia - CNA

He compared AI development to driving a high-performance car: powerful engines require “good brakes,” i.e., rigorous testing, clear limits on autonomous systems, and mechanisms for human intervention. The minister cited three principal risk vectors: loss of control over autonomous systems, misuse by rogue actors to create weapons of mass destruction, and pervasive socio-economic disruption.

At UN, Singapore warns AI safeguards must keep pace as capabilities race ahead

• AI risk categories – loss of control, weaponisation, and socio-economic disruption (Balakrishnan).

At UN, Singapore warns AI safeguards must keep pace as capabilities race ahead

• Strait of Hormuz oil flow – roughly 20 % of worldwide oil shipments pass through the strait (Balakrishnan).

Channel News Asia - CNA

• International support – more than 20 countries and the European Union signed a declaration on the sidelines of the UNGA calling for binding government oversight of frontier AI models (Balakrishnan).

Singapore proposes a UN framework convention on AI safety

Balakrishnan’s address emphasized three interlinked priorities:

Singapore proposes a UN framework convention on AI safety

1.

Nepal PM's impassioned UNGA speech

2. Institutional analogues – He cited the International Telecommunications Union (ITU) and the International Atomic Energy Agency (IAEA) as models for bodies that could set technical standards and oversee safe AI deployment.

Singapore proposes a UN framework convention on AI safety

3.

Trump and Xi strike cordial tone at summit. And, Netanyahu gives defiant UNGA speech

Turkish President Recep Erdogan made a parallel appeal for a common international legal framework for AI during his own UNGA address, reinforcing the multilateral momentum behind AI governance.

Singapore proposes a UN framework convention on AI safety

• “We must collectively formulate new rules and be prepared to create new institutions, if necessary, for new technologies,” — Vivian Balakrishnan, Singapore’s foreign minister (q1)

Singapore proposes a UN framework convention on AI safety

• “Because China missed it, China underwent a century plus of humiliation and unequal treaties. This time, China has absolutely no intention of missing it again,” — Vivian Balakrishnan, Singapore’s foreign minister (q2)

Singapore proposes a UN framework convention on AI safety

• “The largest economies depend on the open sea lanes as much as the smallest, and we in Singapore hold ourselves to the same standard,” — Vivian Balakrishnan, Singapore’s foreign minister (q3)

Singapore proposes a UN framework convention on AI safety

Balakrishnan indicated that discussions on a UN Framework Convention on AI Safeguards are now “on the table,” and that Singapore will continue to engage in multilateral forums to shape the architecture of any future treaty.

Singapore proposes a UN framework convention on AI safety

The speech marked a clear shift from voluntary corporate self-regulation toward binding international oversight of frontier AI models, reflecting growing consensus that AI’s rapid advancement demands coordinated global governance.

Singapore proposes a UN framework convention on AI safety