Singapore’s Infocomm Media Development Authority (IMDA) announced on March 10 that it had rolled out a new model framework for the governance of agentic artificial intelligence. The initiative is part of the city‑state’s broader strategy to keep pace with rapid advances in generative AI while safeguarding public trust.
Key Pillars of the Framework
The framework introduces five core pillars: Transparency, Accountability, Safety and Reliability, Human‑in‑the‑Loop and Fairness & Bias Mitigation. Each pillar is accompanied by a set of principles, best‑practice guidelines and technical checklists that organisations can adopt to demonstrate compliance.
Stakeholder Engagement
IMDA consulted with industry groups, academia, civil society and the private sector over a six‑month period. “We listened closely to the concerns of businesses and citizens alike,” said Dr Evelyn Tan, IMDA’s Director of AI & Emerging Tech.
“The framework is designed not as a regulatory burden but as a practical roadmap for responsible innovation.”
Implications for Business and Public Services
Under the new rules, companies that develop or deploy autonomous AI systems will be required to conduct risk assessments, maintain audit trails and provide mechanisms for user override. Government agencies are expected to adopt the framework as part of their procurement guidelines by 2027.
International Context
The initiative follows similar moves in Europe, the United States and China, positioning Singapore as a regional hub for trustworthy AI. Analysts predict that firms seeking to expand into Southeast Asia will view the framework as a key factor in market entry decisions.