Koon Heng Teow

TEOW-AGL: A Pre-Execution Governance Layer for Agentic AI Systems

2026 · Open MIND

Agentic artificial intelligence systems can now execute actions in the external world before human operators or governance layers have an opportunity to intervene. Existing safety approaches — prompt guardrails, output filters, permission lists, and post-hoc …