Chapter 5: The AAA Licence

Previously in the book

Chapter 01: The F1 Revolution: Formula 1 shows why performance comes from human-machine systems, not technology or heroic individuals alone.

Chapter 02: The Co-Intelligent Vision: A new organisation begins when we design for humans with AI, rather than humans versus AI.

Chapter 03: The 9Q Grid: Map AI agency against the nature of work before deciding where to deploy intelligence.

Chapter 04: Strategic Shapes: The pattern of AI initiatives across a portfolio reveals whether strategy is deliberate or drifting.

This week's chapter: The AAA Licence

Part 1 gave us a map. We explored what changes when humans work with AI, how the 9Q Grid positions different kinds of work, and how strategic shapes emerge across an organisation’s AI portfolio. But a position on the map is not permission to race there.

Oscar Piastri earned his way through the junior categories before reaching Formula 1. Max Verstappen took an exceptional route. Organisations pursuing AI autonomy often behave as if the exception is a repeatable operating model. It isn’t. Ambition and software spend cannot substitute for capability.

The AAA Licence sets out three ways to allocate authority: ASSIST, where people direct, check and decide; AUGMENT, where people and AI work through decisions together with human accountability; and ADAPT, where AI can act within explicit limits while humans monitor and intervene. These are not universal stages every task must climb. Each initiative needs the authority appropriate to its work and the proof to support it.

The chapter brings this to life through the Williams Paradox, Ferrari Gap and customer-service examples. The key test is simple: does the authority delegated to the system exceed the evidence of our ability to govern it? If yes, narrow the mandate or build the missing evidence. You cannot buy an ADAPT licence with ASSIST maturity.

Read Chapter 5: The AAA Licence

What is the biggest gap you see between organisational AI ambition and readiness?