Revised framework supports generative AI use while HSA wins top WHO medical device recognition
Singapore is tightening its grip on healthcare regulation while opening more room for innovation, updating its artificial intelligence framework to support faster adoption of new tools and future AI-developed drugs.
Singapore Updates Its Healthcare AI Framework
Health Minister Ong Ye Kung said on March 10 that Singapore has revised its healthcare AI guidelines to reflect newer developments such as generative AI, while still safeguarding safety and quality. The revised framework was co-developed by the Ministry of Health and the Health Sciences Authority to better support innovation in healthcare settings.
Ong said the changes are intended to help AI augment the healthcare workforce and eventually allow new therapies, including AI-developed drugs, to reach patients faster. He added that Singapore wants to support innovation without compromising regulatory rigor.
Sandboxes Will Test AI in Real Clinical Settings
A key part of the revised approach is the use of regulatory sandboxes, which will allow AI solutions to be evaluated in real-world healthcare environments. Ong said this helps ensure AI tools are trained and tested using quality, real-life data rather than only controlled lab conditions.
This direction builds on Singapore’s broader AI-in-healthcare regulatory work, including HSA’s consultation on exemptions for selected public-sector AI software as a medical device. The goal is to encourage practical deployment while preserving accountability and patient safety.
HSA Wins WHO’s Highest Device Regulatory Rating
Ong also announced that HSA has become the first national regulatory authority to attain the World Health Organization’s highest maturity level for medical device regulation, known as Maturity Level 4. That recognition allows HSA to serve as a global reference point for other regulators.
He said the recognition strengthens Singapore’s credibility as a regional and global regulatory node. Ong added that a number of jurisdictions already reference HSA approvals to speed up their own medical device or pharmaceutical processes.
AI-Developed Drugs Could Be Next
Ong said HSA has not yet received any application for an AI-developed drug, but would welcome such submissions in the future. He noted that AI is beginning to reshape drug development, including through the use of simulated laboratory data that may reduce reliance on some costly early-stage trials.
He said HSA will take a technology-neutral regulatory approach, applying the same level of scrutiny to AI-developed drugs as it does to conventional medicines. That is meant to ensure innovation moves forward without weakening standards for efficacy or safety.
Biomedical Growth Is Part of the Strategy
Ong said Singapore also wants HSA’s regulatory strength to reinforce the country’s biomedical competitiveness. He outlined plans to better align product development, regulatory approval, health technology assessment, and clinical application in priority disease areas such as cardiovascular disease, diabetes, and metabolic disorders.
HSA chief executive Raymond Chua said the agency will work with universities, hospitals, academic medical centers, and A*STAR to strengthen Singapore’s biomedical innovation pipeline and grow clinical trial capabilities. Taken together, these moves are meant to position Singapore as more than a small domestic market, but as a gateway to wider regional access.
Singapore’s revised healthcare AI framework shows it is trying to move faster on medical innovation without loosening regulatory discipline. For Singaporeans, that could mean earlier access to safer AI tools and faster therapeutic development. For Indonesians and others in the region, Singapore’s growing status as a trusted health regulator may also shape how new medical technologies, devices, and drugs enter Southeast Asian markets.
Sources: Straits Times (2026) , WHO (2026)
Keywords: Ong Ye Kung, HSA WHO Level 4, Healthcare AI Framework, Generative AI Healthcare, Singapore Biomedical Sector
