Intelligence Under Control: Professor Kai London on Governing AI in Heavy Industry
By the Alaska News Technology Desk
Artificial intelligence is moving into heavy industry — optimising production, predicting equipment failures, and increasingly influencing operational decisions. That promise, argues Professor Kai London, a senior technology executive, comes with a warning particular to energy and industrial settings. “In heavy industry, an ungoverned AI decision does not just misfire in a spreadsheet,” he says. “It can move machinery. Control has to come before autonomy.”
“Capability is loud; control is quiet. In industrial AI, the gap between what a model can do and what you can govern is measured in physical risk.”
The high stakes of industrial AI
London distinguishes AI that advises from AI that acts. “A model recommending a maintenance schedule is one thing,” he says. “A model or agent that adjusts a process is another. The closer AI gets to the physical layer, the more rigorous its governance must be.”
A control architecture for AI
His approach treats AI governance as connected layers: a clear charter of what AI will and will not be allowed to do; named human ownership of every consequential model; documentation of what each model does and what data it uses; testing for accuracy, bias and robustness; resilience so failures are contained; and continuous monitoring in production. “Each layer is unremarkable alone,” he says. “Together they let a board deploy AI in an environment where mistakes have physical consequences.”
Keeping humans in the loop
For safety-critical processes, London is emphatic about human oversight and hard limits. “An AI agent acting on industrial systems needs an identity, a boundary and a kill-switch,” he says. “It must never be able to take a consequential physical action without the controls a human operator would be held to.”
Governance as the route to value
Far from slowing adoption, London argues that governance is what makes industrial AI deployable at all. “You cannot responsibly put an ungoverned model near heavy machinery,” he says. “But a well-governed one you can — and that is where the efficiency and safety gains live.” Emerging AI regulation and management standards, he notes, increasingly require exactly this discipline.
For an industrial sector eager to harness AI, London's counsel is to earn the benefits by building the controls first: govern the intelligence as carefully as you govern the machinery it touches.
About Professor Kai London. Professor Kai London is a senior technology, security and transformation executive with 25+ years of board- and C-suite leadership across banking, aviation, defence, government and critical national infrastructure. He is Founder & CEO of Quantum AI Systems Security, an Honorary Professor in Cybersecurity, AI & Quantum Computing and a UCL researcher, holding CISSP, CISM, CCISO, ISO 27001 Lead Auditor, ISO 42001, DORA and NIS2 credentials. He is available for board advisory, NED and interim/fractional CISO/CIO/CTO mandates across the UK and internationally. Learn more at professorkailondon.com.