Governing Machine and AI-Agent Identities: Professor Kai London on the Worst-Controlled Class
By the Resolute Islands News Technology Desk Professor Kai London — board advisor & interim CISO/CIO/CTO. Credit: professorkailondon.com Amid all the attention on human passwords and phishing, Professor Kai London , a senior CISO, points to a quieter and faster-growing danger. “The fastest-growing and worst-controlled class of identity in almost every organisation is not human,” he says. “It is machines — service accounts, API keys, automation — and now AI agents that can act on their own.” “Every AI agent you deploy is a login with real reach. It needs an identity, a boundary, monitoring and a kill-switch — exactly like a privileged human user, only faster and at greater scale.” The invisible workforce London notes that machine identities typically outnumber human ones many times over, and are often granted broad, standing privileges that nobody reviews. “They were set up to make systems work, not to be governed,” he says. “That is precisely the problem.” AI agents change th...