Singapore has unveiled key insights from its Global AI Assurance Pilot, a pioneering initiative aimed at establishing norms and best practices for the technical testing of Generative AI (GenAI) applications. Announced at the ATxSummit 2025 by Senior Minister of State for Digital Development and Information, Tan Kiat How, the initiative positions Singapore at the forefront of AI safety and responsible adoption.
The pilot, launched in February 2025 by the AI Verify Foundation and the Infocomm Media Development Authority (IMDA), paired 16 specialist AI testers with 17 GenAI deployers across 10 industries, including finance, healthcare, and public sectors. A significant finding was the context-dependent nature of GenAI risks, which vary by industry, use case, and culture. The pilot recommends involving subject matter experts throughout the application lifecycle to mitigate these risks.
In conjunction with the pilot, IMDA announced the development of the world’s first Testing Starter Kit for GenAI applications. This kit provides practical guidance for businesses on testing GenAI applications, addressing common risks such as hallucination, undesirable content, and data disclosure. IMDA is seeking industry feedback on the kit, which will evolve to address emerging risks.
Additionally, AI Singapore (AISG) will partner with the UN Development Programme to enhance AI literacy in six pilot countries, expanding the successful AI for Good programme. The AI Student Developer Conference, led by AISG, also highlighted Singapore’s commitment to fostering AI talent, featuring a regional challenge with students from six ASEAN countries.
These initiatives underscore Singapore’s dedication to harnessing AI for public good, promoting economic potential, and enhancing social impact globally.