The WHAT must always have a WHO behind it.

Last week I wrote about platform decisions. This week: when your AI system makes an error, perpetuates a bias, or falls out of compliance, is the WHAT or the WHO responsible?

You step into a lift every day without thinking about it. You are not trusting the lift. You are trusting everything behind it. The engineers who designed it, the company that maintains it, the regulator who certified the building. The moment you cannot answer who is behind it, you take the stairs.

We have always trusted inanimate objects. The lift, the plane, the bridge, the medication. We do it without hesitation because behind every one of them there is a visible, traceable chain of human accountability. Something goes wrong, and we know who to call. The object is the what. The accountability belongs to the who.

AI is no different in principle. But most organisations are deploying it without ever answering that question. When the system makes an error, who is responsible? When it perpetuates a bias, who corrects it? When the data it was trained on becomes outdated, who retrains it? When a regulator asks whether it remains compliant, who answers?

These are not edge cases. They are the operating conditions of any AI system in a real organisation. And they all point back to the same place. A human. A named, accountable human WHO owns the WHAT.

Designing an AI strategy without answering that question is not just a governance gap. It is increasingly a legal one. Regulation across major markets is moving in one direction: organisations will be required to demonstrate that a responsible human actor sits behind every consequential AI decision. The what must have a who, and that who must be findable.

AI innovation is not just a technology question. It never has been. If you would like to discuss how your organisation is thinking about this, reach out and lets discuss.

Where have you seen the who disappear behind the what in an AI deployment?