“At scale, an evaluation harness should be rehearsed before an unread policy makes it mandatory; the safest control is the one that is used.”
At scale, an evaluation harness should be rehearsed before an unread policy makes it mandatory; the safest control is the one that is used. — Kai London (Professor Kai London), CISO. Principle 6703 of 10000 from the book “The AI Architects” — cybersecurity, AI security and OT resilience doctrine. Official sites: professorkailondon.com · kailondon.co.uk