Singapore announced comprehensive artificial intelligence (AI) governance initiatives in Paris on February 11, 2025, during the AI Action Summit (AIAS). The measures, outlined by Minister for Digital Development and Information Josephine Teo, aim to increase the safety of AI applications for Singaporean and global users, responding to the cross-border nature of AI products and services.
The announced initiatives are built on three main pillars: the Global AI Assurance Pilot, which aims to establish global best practices for technical testing of generative AI (GenAI) applications; the Testing Report jointly prepared with Japan, which examines the safety protections of large language models (LLMs) in 10 different languages (including Cantonese, Farsi, Swahili, and Telugu) and 5 harm categories; and the Singapore AI Safety Red Teaming Challenge Evaluation Report, during which more than 350 participants from 9 Asia-Pacific countries tested the cultural bias of 4 LLMs (Aya, Claude, Llama, SEA-LION) in non-English environments.
The initiatives build on the results of the Bletchley Park summit in November 2023 and the Seoul summit in May 2024, demonstrating Singapore's active role in shaping global AI standards. As part of the program, Singapore joined the "Inclusive and Sustainable Artificial Intelligence for People and the Planet" leadership declaration, reinforcing its commitment to transparent, inclusive, and accountable AI development while taking concrete steps to develop the third-party AI assessment market under the leadership of the AI Verify Foundation and IMDA.
Sources:
1.
Alita Sharon
2.
Singapore is introducing new AI governance policies for safe and ethical technology use.
3.
