Open access
2025
Human Oversight for Autonomous AI Agents: A Governance Framework for Healthcare, Finance, Energy, and Public Infrastructure
The findings show that meaningful oversight is not a single human approval step and is a lifecycle capability that combines bounded autonomy, evidence-based escalation, stop authority, continuous validation, audit records, and institutional learning.
Aaron K Montgomery, Hannah E Gallagher, Derrick L Mercer
· International Journal of Eng... · 0 citations