Designing the AI Op Model

Who owns AI here?

The Chief Data Officer thinks they do, and so does the CTO. The CFO owns the funding and General Managers are busy using that to demonstrate their ability to integrate AI into their business functions. Ironically, the organisation chart doesn't define it; it only includes humans.

Nobody should own AI. But somebody must own the decisions it makes.

What's at stake

Only 1 in 8 CEOs report that AI has influenced both revenue and costs positively (PwC) and about 7 out of 8 organisational leaders report that their organisations are not very well prepared to adopt AI in day-to-day operations (McKinsey). Those numbers point to something more interesting than an adoption problem. If this were simply a matter of buying better technology, the problem might have been solved by now. Maybe it is how organisations are adopting it?

Early digital programs often digitised existing processes before organisations learned to redesign the customer journey around what digital made possible. AI appears to be following a similar pattern. We are making existing tasks faster before asking the more consequential question: if intelligence becomes cheap, scalable and increasingly autonomous, how should the process work in the first place?

The lesson to be learnt from past digital transformations is that by just applying technology to the same organisational design and processes, only part of the value can be captured. Organisations that took a holistic approach and considered how technology could help redesign the experience rather than just applying it to existing processes captured more of the opportunity.

Yet despite the opportunity to redesign work, most leaders still expect AI to function primarily as a support tool over the next two years.

How to start

If you make the right decisions, in the right order, it might not be as hard as you think.

1. Where will AI create disproportionate value? Which three or four customer journeys, operational domains or decision systems matter enough that redesigning them could materially affect revenue, margin, risk, working capital or customer experience?

Don't just think about the core priorities, revisit past opportunities that have been discarded or considered too hard. Not just improvements but new (or forgotten) opportunities. AI changes the economics in many cases.

2. Who owns the outcome, the capability and the risk? For every priority AI-enabled domain, be clear about five different owners: the outcome, the process, the data, the enabling capability and the risk. They may not be the same person.

3. How should the work be redesigned? Break the priority workflow into decisions and activities. Which should remain human-led? Which should be AI-assisted? Which can become AI-led with human supervision? And which could eventually operate autonomously?

The answer isn't determined by technical capability alone. It reflects regulatory obligations, risk appetite, customer expectations, economics and your value proposition.

From the 10th December 2026 many organisations have new responsibilities when using personal information to automate decisions. Review the Privacy Act transparency obligations now.

For affected organisations, understanding where decisions occur is becoming not just good operating-model design, but part of regulatory readiness.

4. Funding & Measurement. Funding AI as projects and it dies as pilots. Fund it as a technology platform and it loses all business accountability. But, funded as P&L owned capability on a shared platform, focused on measuring outcomes and you've established something that will scale.

Finding the right balance between centralised capability and distributed P&L aligned funding can be challenging. Review investment cycles and line these up as much as possible, develop a roadmap and fund the gaps centrally to avoid fragmentation.

5. Role Design. AI changes the economics of team design. Agents are increasingly being deployed to speed up analysis, research and automate tasks. This is elevating human capabilities like judgement, domain expertise, exception handling, relationship management and accountability. This is likely to produce smaller, leaner teams but doesn't negate the need to develop the experience required to achieve those higher-order tasks in the future.

Where to start

Before an organisation can delegate work to machines, it has to understand what it is delegating.

This is where things get harder. Most organisations are far less explicit about how they actually operate than they think they are. Policies describe the normal case, but exceptions live in people's heads. Decision rights are often implicit. Process logic is embedded in systems, and experienced employees have learned how to navigate the gaps.

Humans can work around that ambiguity. Reliable autonomous agents need much more of it made explicit.

This isn't necessarily a model problem. It is increasingly an organisational knowledge problem.

Treat and fund knowledge management like infrastructure.