Singapore – Organisations often favour workers who can work in a team, yet remain independent and reliable when assigned specific tasks or projects.
Today’s modern age brings a new era of autonomous workers who are constantly pursuing action items in the background, accessing various tools and resources, while enabling human teams to focus on the growth and development of company operations.
These independent workers are: AI agents.
As the popularity of AI agents and autonomous workflows surges across global markets, especially in the Asia Pacific, questions begin to arise surrounding how organisations are keeping intelligent agents accountable for their actions with human oversight.
Following this, Singapore’s Infocomm Media Development Authority (IMDA) officially launched the world’s first Model AI Governance Framework (MGF) for agentic AI. This marks a significant milestone for the future deployment of agentic systems across the region.
With Singapore being the springboard for AI innovation and adoption, the MGF provides organisations with a structured overview of the risks and emerging best practices for managing the deployment of agentic AI operations.
More importantly, the MGF is not just a stationary framework—but rather is living and breathing—adapting to new advancements in AI technologies, with the support of feedback from public and private entities.
The framework covers four key considerations to help organisations with the deployment of autonomous agents: Assess and bound the risks upfront, make humans meaningfully accountable, implement technical controls and processes, and enable end-user responsibility.
With the MGF officially available, how will this framework reshape the way organisations manage accountability and risk for AI agents?
Exploring this, UpTech Media spoke with Kris Day, senior vice president and general manager for APJ at SentinelOne and Paul Carvouni, senior vice president and general manager for ASEAN at Salesforce, with comments from Takanori Nishiyama, SVP APAC & Japan Country Manager at Keeper Security.
Autonomous accountability
It is no secret that AI agents have entered the workforce as complementary, autonomous workers to help remove friction across company-wide operations. While this is intended to allow human teams to focus on priority tasks, it leaves the AI agents unsupervised.
This is where Singapore’s MGF comes into play, equipping organisations to deliver accountability and transparency for autonomous agents. Kris mentioned how the MGF recognises that agentic AI is rapidly moving into real-world use and focuses on the safe and responsible deployment of autonomous agents.
“The framework’s emphasis on bounding an agent’s access to tools and data, enforcing human checkpoints for high‑impact actions, and applying technical controls across the agent lifecycle mirrors how leading security teams want to operationalise autonomous SOCs,” Kris commented.
He explained how in the world of cybersecurity, the MGF reinforces a ‘security for AI and AI security’ approach that empowers real-time detection, investigation, and response under clear guardrails for testing and transparency. The framework pushes organisations across all industries to be accountable for the actions of their autonomous agents.
At the same time, Takanori echoed similar sentiments and addressed how the MGF allows organisations to explicitly address risks such as unauthorised actions, data misuse, and systemic disruptions for autonomous agents to apply best-in-class principles when it comes to enterprise identity governance and AI oversight.
“They set the benchmark for responsibly managing autonomous systems, protecting sensitive data and maintaining operational resilience, which other countries in the APAC region can emulate,” Takanori added.
Managing risks and accountability for AI agents is a constant balancing between human intervention and autonomous operations. As Paul added, the MGF provides practical guidelines for responsible agentic AI deployment where companies need to scale safety, while emphasising human oversight.
“Singapore has shown that effective governance doesn’t slow down innovation— it’s what enables it,” Paul commented.
Responsible agentic scalability
The MGF was designed by both private and public entities as a framework not just for the deployment of AI agents, but for the entire lifecycle of their autonomous workflows as organisations scale and expand.
Takanori highlighted how the framework encourages AI agents to be treated as employees within a company to safely monitor their actions and mitigate unnecessary risks.
“The framework underscores this approach: assigning each AI agent a verifiable identity, enforcing task-specific, time-bound permissions and ensuring human accountability at every stage. These measures reflect the standards necessary for safely deploying AI at scale, where visibility, control and auditability are non-negotiable,” he stated.
Meanwhile, Kris added to this by showcasing how the MGF gives policymakers, boards, and CISOs a shared playbook for understanding where autonomy is appropriate, when human oversight is required, and how agentic AI performance should be monitored over time.
“Well-designed guardrails accelerate adoption. By positioning the framework as sector-agnostic and evolving, Singapore provides regulatory clarity that encourages responsible investment rather than hesitation,” Kris shared.
With the MGF placing Singapore to lead governance for agnetic AI, it also opens a gateway for other nations across APAC to begin implementing their own framework for proper governance, accountability, and human oversight towards the future of AI technologies.
“By being risk-based and tailored to the specific nature of agentic systems, rather than applying a one-size-fits-all model, it creates space for businesses to deploy responsibly while still moving at pace,” Paul added.
“That balance––guardrails that give organisations the confidence to go faster, not slower––is what every market in the region should be working toward.”
Vietnam has already taken the first steps by introducing the nation’s own comprehensive legislative law governing the deployment of AI.
****
The digital era has ushered in a new wave of autonomous and independent workers via AI agents.
While helpful, necessary oversight to mitigate risks and maintain accountability is essential, and Singapore’s MGF provides the framework to help pave the ever-evolving agentic future.
