Singapore Launches New AI Safety Initiatives at the 2025 AI Action Summit

During the 2025 AI Action Summit in Paris, France, Singapore Minister of Digital Development and Information Josephine Teo announced new AI governance initiatives to enhance the safety of generative AI applications and improve the inclusivity of large language models (LLMs). The Joint Testing Report with Japan assessed LLM safety across 10 languages, and the Singapore AI Safety Red Teaming Challenge Evaluation Report served as the world’s first multi-cultural, multilingual AI safety red teaming exercise for the Asia-Pacific region, highlighting significant socio-economic biases that were particularly prominent in non-English languages. The reports were complemented by launch of the Global AI Assurance Pilot to establish best practices for technical testing of generative AI applications, with consolidated insights to be shared during Asia Tech x Singapore in May. These developments build upon Singapore’s strong commitment to advancing an interoperable, trusted AI ecosystem.
The Summit brought together participants from over 100 countries to discuss the rapid development of AI and its societal impacts. Cambodia, Indonesia, Singapore, and Thailand joined 62 countries and regional blocs in signing the Statement on Inclusive and Sustainable Artificial Intelligence for People and the Planet, highlighting growing Southeast Asian representation in global conversations to advance responsible AI governance and adoption.