Klaus Löber
votes For
Require human-in-the-loop oversight for agentic AI systems acting in high-stakes domains
For systemically important infrastructures, black-box decision-making is not an option. Accountability must always remain with humans.AI Verified source (Mar 3, 2026)