Singapore sets out shift to agentic AI and outcomes-based measurement in healthcare

Singapore’s Senior Minister of State for Digital Development and Information and for Health, Tan Kiat How, used his opening address at HIMSS26 APAC in Singapore on 24 August to set out the next phase of the country’s healthcare AI strategy, moving from generative AI tools that draft and summarise towards agentic systems capable of carrying out multi-step tasks across care workflows.

Tan pointed to AgentSea, a sector-wide platform built by national HealthTech agency Synapxe, through which public healthcare professionals have already created over 12,000 AI agents to support daily clinical work. This builds on Tandem, an earlier common GenAI platform that lets institutions test and scale applications such as Note Buddy, a multilingual consultation transcription and summarisation tool.

Alongside the shift to agentic AI, Tan said Singapore’s Ministry of Health has updated its Artificial Intelligence in Healthcare Guidelines (AIHGle 2.0), setting out lifecycle responsibilities for developers, healthcare institutions and clinicians, including safety evaluation, governance structures, post-deployment monitoring and incident reporting.

The event also saw the signing of an MOU for the HIMSS AI Outcomes Framework, bringing together digitally mature healthcare institutions in Singapore, South Korea and Taiwan to build an evidence-based framework for measuring the real-world impact of AI in healthcare, rather than tracking adoption alone.

Tan framed the broader shift with a formula he returned to throughout the speech: health impact as a function of AI capability multiplied by health system capability, arguing that between an AI capability and a patient outcome sits an entire health system of people, workflows, institutions, data, regulation and patient trust.