“A human-in-the-loop gate must be pausable, explainable, and controllable — because an agent you cannot pause is an agent you do not control.”

— Professor Kai London, CISO.

Quote card by Professor Kai London reading 'A human-in-the-loop gate must be pausable, explainable, and controllable — because an agent you cannot pause is an agent you do not control.'. From the book The AI Control Architecture.
Quote card by Professor Kai London reading 'A human-in-the-loop gate must be pausable, explainable, and controllable — because an agent you cannot pause is an agent you do not control.'. From the book The AI Control Architecture.

A human-in-the-loop gate must be pausable, explainable, and controllable — because an agent you cannot pause is an agent you do not control. — Kai London (Professor Kai London), CISO. Principle 1068 of 10000 from the book “The AI Control Architecture” — cybersecurity, AI security and OT resilience doctrine. Official sites: professorkailondon.com · kailondon.co.uk

Related: Principle 3568 Principle 6068 Principle 8568

View in gallery (page 11) · About this book