AI Governance Expands as Autonomous Systems Enter the Physical World
AI Governance Expands as Autonomous Systems Enter the Physical World
Artificial intelligence is increasingly moving beyond software applications and into the real world through robots, autonomous vehicles, delivery systems, and industrial machines. As these technologies become more common, experts and policymakers are raising questions about whether current AI regulations are sufficient to manage systems that operate in physical environments.
Unlike traditional AI concerns such as misinformation, bias, and harmful online content, embodied AI introduces risks that can directly impact people, infrastructure, and public safety. Failures in these systems could affect transportation networks, warehouses, power grids, and other critical services.
In response to these challenges, Singapore's Infocomm Media Development Authority recently released Version 1.5 of its Model AI Governance Framework for Agentic AI. The framework provides guidance for organizations deploying AI agents capable of making decisions, using external tools, and carrying out tasks with limited human intervention.
The framework recommends strict access controls, continuous monitoring, human approval checkpoints, and emergency shutdown mechanisms for AI systems operating in real-world environments. It also emphasizes that organizations remain accountable for the actions of their AI agents, even when those systems function autonomously.
During a recent AI summit in Singapore, experts highlighted the growing need for governance models focused on deployment and ongoing monitoring rather than relying solely on pre-release testing. Researchers warned that risks in digital systems can become significantly more serious once AI is integrated into physical machines such as drones, autonomous vehicles, and smart infrastructure.
Technology companies are already putting these principles into practice. Ride-hailing giant Grab is testing autonomous vehicles and delivery robots in Singapore. The company says extensive simulations, real-world testing, and continuous monitoring are essential before scaling deployments.
Meanwhile, countries across Asia are accelerating efforts to develop standards for embodied AI. China is expanding industrial robotics through government-backed initiatives, while Japan is investing in robotics datasets, safety research, and governance frameworks to support future AI-powered machines.
Businesses are also increasing the use of agentic AI in regulated industries. Financial institutions such as JPMorgan Chase are deploying AI tools to assist employees with research, document analysis, and client engagement while maintaining human oversight for critical decisions.
The adoption of AI-powered robots is also growing in Japan. A recent survey found that nearly one-third of Japanese companies are already using, planning to use, or considering AI-driven robotics, particularly in manufacturing and hazardous work environments.
Experts say the challenge moving forward will be finding the right balance between innovation and safety. As autonomous AI systems gain greater responsibility in the physical world, governments and companies will need governance frameworks that ensure these technologies remain reliable, accountable, and safe for public use.
