Singapore Launches New AI Governance Initiatives at Global Summit

Singapore has announced three new initiatives to enhance AI governance and safety, unveiled by Minister for Digital Development and Information, Josephine Teo, at the AI Action Summit (AIAS) in Paris. These initiatives aim to address the transboundary nature of AI products and services, benefiting both Singaporeans and the global community.

The first initiative, the Global AI Assurance Pilot, focuses on establishing best practices for the technical testing of generative AI (GenAI) applications. This pilot, launched by the AI Verify Foundation and the Infocomm Media Development Authority (IMDA), will convene AI assurance and testing vendors with firms deploying real-life GenAI applications. It aims to shape future AI assurance standards and contribute to the growth of local and international third-party AI assurance markets.

The second initiative is a Joint Testing Report with Japan, conducted under the AI Safety Institute (AISI) Network. This collaboration assesses Large Language Models (LLMs) across different linguistic environments, focusing on evaluating guardrails for LLM safety in non-English settings. The report supports the development of evaluation capabilities and methodological standards for safer AI models in diverse languages.

The third initiative is the Singapore AI Safety Red Teaming Challenge Evaluation Report 2025. This report evaluates how LLM safeguards perform in diverse cultural and linguistic settings within the Asia-Pacific region. The challenge involved over 50 participants from nine Asia-Pacific countries testing four LLMs for cultural bias and stereotypes in non-English languages.

These initiatives underscore Singapore's proactive stance in enhancing AI safety through international collaboration and rigorous testing. By fostering partnerships and developing robust evaluation frameworks, Singapore is contributing to a safer and more accountable AI landscape for the global community.