Singapore has joined the European Union and 19 other nations in advocating for enhanced global oversight of frontier artificial intelligence (AI) models. This initiative was formalized through a joint declaration adopted during the United Nations General Assembly in New York, aiming to balance the benefits of AI with the risks posed by increasingly powerful systems.
Prime Minister Lawrence Wong highlighted the necessity of international cooperation in managing AI risks that transcend national boundaries. Emphasizing that AI should always remain under human control, Wong underscored Singapore’s commitment to international efforts that ensure AI advancements contribute positively to society.
The declaration identifies AI’s potential to improve lives, advance scientific research, and bolster economic development while warning of the serious safety and security risks that could arise if frontier AI models are not properly managed. It calls for the development of an international institution dedicated to frontier AI, which could establish standards and coordinate responses when AI systems reach significant capability levels.
The signatories urge UN member states to build on existing international mechanisms and explore the creation of common AI standards, promote transparency, and develop shared methods for reporting serious AI safety incidents. Additionally, the declaration emphasizes the need for increased access to scientific expertise and reliable AI evaluation capabilities worldwide.
AI companies are encouraged to enhance their safety practices by thoroughly testing systems before deployment and engaging independent evaluations to identify potential risks. Despite the broad support, notable absences from the endorsement include the United States and China, two leading powers in the AI field. Singapore stands as the sole Southeast Asian nation among the signatories, with the declaration remaining open for additional countries to join the initiative for stronger AI safeguards.