Singapore’s Infocomm Media Development Authority (IMDA), its digital and media regulator, has released a starter kit to help organisations test AI agents for safety and reliability, and has set up a task force on cyber threats from the most advanced AI. CNA and The Straits Times reported the announcements, made by IMDA at an industry event on 14 October 2026.
The kit is voluntary guidance for organisations, not a law or regulation.
Testing how an agent gets there, not just where it ends up
AI agents are AI systems that can act on a user’s behalf, use tools, access other systems and sensitive data, and make changes. “When we give AI more agency, we need greater assurance over how that agency is exercised,” IMDA chief executive Ng Cher Pong said, according to CNA.
According to CNA, the kit sets out four steps: identifying risks such as erroneous or unauthorised actions, designing realistic tests, analysing logs, and continuously monitoring and re-testing systems. It calls for testing an agent’s “entire trajectory, including its reasoning and actions”, rather than only its final output.
CNA reported that the kit draws on consultations with more than 20 industry deployers and testers and on work by the Singapore AI Safety Institute. The Straits Times said those consulted included Amazon, Google, Microsoft and Singtel, and that the kit builds on a governance framework for agentic AI that Singapore launched in January.
“As AI evolves rapidly, testing methods have to catch up too. The Starter Kit for Testing Agentic AI Systems for Safety and Reliability is therefore a living document, which we will continue to update and refine, and we welcome your feedback along the way,” Ng said, according to CNA.
A task force on frontier-AI cyber risks
IMDA also formed the Frontier AI Resilience and Security Taskforce (FAIRS), made up of more than 20 experts and industry leaders from telecommunications, media, cybersecurity and AI, CNA reported. It is expected to publish a report with practical guidelines in the first half of 2027.
According to The Straits Times, the report will focus on three priorities: strengthening responses to AI-accelerated threats, improving resilience against emerging AI risks, and keeping up with a changing cyber threat landscape as frontier AI develops.
“Frontier AI is changing the rules of the game and we must adapt accordingly,” Ng said, according to CNA. He added that some risks cannot be managed just by testing individual AI applications.
What this means
- Agents need new kinds of tests. Singapore wants firms to check every step an agent takes.
- It is guidance, not law. The kit is voluntary and will be updated over time.
- Frontier AI is a cyber issue. The task force will look at threats beyond single apps.
Sources
- CNA, “Singapore forms task force to strengthen defences against AI-enabled cyber threats”, 14 October 2026. https://www.channelnewsasia.com/singapore/singapore-task-force-ai-related-cyber-threats-6457921
- The Straits Times, “Singapore launches new toolkit to help firms test AI agents for safety and reliability”, 14 October 2026. https://www.straitstimes.com/singapore/singapore-launches-new-tool-kit-to-help-firms-test-ai-agents-for-safety-and-reliability
