Issue 4 The Operating Issue

In this issue

The Agent Reckoning — Why 2026 is the year AI pays for itself

The New Operating Model — Nine CIOs on what they changed

Field Notes: Running Inference at National Scale — A technical and operational case study

The Assurance Questionnaire Boards Are Using — The due diligence framework every board needs

Interview: A Chief Data Officer on Killing Four Use Cases — Why saying no is the most important skill in AI

The era of experimentation is over.

For the past two years, enterprises have been piloting AI. They have been running proofs of concept. They have been exploring use cases. They have been learning.

Now, they are being asked to move from experimentation to operations.

The Operating Reality

This issue of AI Works is about what it really takes to run AI at scale. Not as a science project. Not as a technology demonstration. As a core business function.

This shift is not trivial. It requires changes to funding models. It requires changes to organisational structures. It requires changes to governance. It requires changes to culture.

And it requires a clear-eyed understanding of the unit economics of AI — what it costs, what it delivers, and whether it pays.

Funding AI

One of the biggest questions facing enterprise leaders is how to fund AI initiatives. Should AI be funded centrally, like IT? Should it be funded by business units, like marketing? Should it have its own budget line?

We examine the emerging models, the trade-offs, and the lessons from organisations that have already made the transition.

Accountability

Who is accountable for AI in your organisation? Is it the CIO? The Chief AI Officer? The Chief Data Officer? The business unit leader?

We explore the different accountability models and the implications for decision-making, risk management, and performance.

The Unit Economics of AI

Perhaps the most important question of all: does AI pay for itself?

We look at the costs — infrastructure, data, talent, governance. We look at the benefits — revenue, efficiency, customer satisfaction. And we look at the metrics that matter for evaluating AI investments over time.

What Leaders Are Doing

This issue is based on extensive reporting. We interviewed nine CIOs and Chief AI Officers about what they have changed in the past 12 months to move from experimentation to operations.

Their answers are surprising, practical, and essential reading for anyone leading AI deployment.

The Bottom Line

The shift to operations is the defining challenge of 2026. Those who get it right will build sustainable competitive advantage. Those who don’t will waste billions.

This issue is your roadmap.

Other Issue

2026-05-26

Issue 3 The Data Issue

Foundations, lineage and the governance work that decides whether anything scales.

2026-03-26

Issue 2 The Risk Issue

Threat models, assurance and what happens when an autonomous system gets it wrong.

2026-01-01

Issue 1 The First Issue

Why a publication for the people accountable for deploying AI — and what we intend to cover.