

A day of sessions on AI in the SOC, OT resilience, fragmenting regulation and machine-speed response.
We build systems that assume the human in the loop has steady attention and reliable judgement. Neither is true, and it is getting less true. Every alert, prompt, exception, and “are you sure?” draws from the same depleting account. By afternoon, the analyst approving a transfer, the engineer waving through a change and the executive clicking an MFA push are all running on the same low battery.
AI does not solve this. It compresses the timeline and raises the stakes of each remaining human decision. The questions we still hand to people – is this normal? do I trust this? should I escalate? – are exactly the ones tiredness destroys first.
You cannot train your way out of the fact that attention runs out. It is a design problem. That’s precisely what I discuss in The Psychology of Information Security.