Agent autonomy raises hard questions about what an AI system may do, how independently it should act, and where human control must stay in the loop. This course builds a practical understanding of autonomy in workplace settings, then connects that to IMDA guidance, structural limits, oversight, testing, monitoring, and end-user responsibilities. It is for people designing, deploying, or supervising autonomous agents who need a clear way to make decisions and spot weak controls before they cause harm.
The course covers workplace definitions of agent autonomy, the difference between decision freedom and action space, and IMDA recommendations for bounding permissions. It also covers structural controls, human approval checkpoints, pre-deployment testing, post-deployment monitoring, and the responsibilities users must understand when working with autonomous agents.