Ethical Constraints and Guardrails in Agent Planning: Embedding Ethical Principles That Override Utility-Maximising Actions
Autonomous and semi-autonomous agents are increasingly used to plan, decide, and act in complex environments. From workflow orchestration systems to AI copilots and decision-support engines, agent planning frameworks often rely on optimisation logic that aims to maximise efficiency, speed, or reward. While such utility-maximising behaviour is technically effective, it can lead to outcomes that conflict […]
