IMDA and Microsoft Team Up on AI Safety in Landmark Singapore Partnership

Singapore's Infocomm Media Development Authority (IMDA) and Microsoft have signed a memorandum of understanding to deepen collaboration on artificial intelligence safety and security, marking the latest major government-Big Tech partnership in the city-state's AI governance push.

The agreement, signed by IMDA deputy chief executive Kiren Kumar and Microsoft's chief responsible AI officer Natasha Crampton, covers three core areas: technical and research collaboration, information sharing, and the development of a policy framework for trusted access to frontier AI models. Under the MOU, IMDA's Singapore AI Safety Institute will work with Microsoft and other government agencies to explore how governments and infrastructure operators can safely access frontier models for security testing.

The partnership also includes joint research into agentic AI — systems that can act autonomously on a user's behalf — alongside the development of evaluation benchmarks for multilingual AI safety. This is particularly relevant for Singapore's multi-ethnic, multilingual context, where AI systems must perform reliably across English, Mandarin, Malay, and Tamil.

The two organisations will jointly develop a white paper examining demand-side needs from government agencies as well as supply-side policy considerations for model providers. Microsoft will also share best practices on governance frameworks and research findings with IMDA and, where appropriate, with Singapore's broader ecosystem partners.

Why it matters for Singapore: This MOU gives Singapore a front-row seat in shaping how frontier AI models are accessed and tested by governments worldwide. As one of the first countries to formalise a trusted-access framework with a major model provider, Singapore strengthens its position as a neutral ground for AI governance — a role that directly supports its ambition to be a global AI hub. The agentic AI research stream is also timely, given Singapore's push to deploy AI agents across its public service.