.png)
Why most global firms are optimising AI at the wrong level and a framework for understanding how they climb
Most firms think AI is a technology investment. The firms creating durable competitive advantage increasingly treat it as an organisational design problem. WPP has created a dedicated global business unit for AI-driven transformation. Ant International has opened a Global Development Centre in Kuala Lumpur, with over half its 1,500 staff in technology roles. IMD has declared that the CIO title is migrating from Chief Information Officer to Chief Intelligence Officer. These are not announcements about tools. They are announcements about organisational architecture.
Organisational architecture, defined here as the configuration of decision rights, talent, capabilities, governance structures, and operating processes that determine how a firm creates competitive advantage, is not a technology decision. It is a strategic one. Yet in most firms, AI strategy is still managed as an IT initiative rather than an architecture question. That misclassification is the root cause of the maturity gap this framework seeks to address.
Most global firms are optimising AI at the wrong level. The competitive advantage has shifted from deployment to architecture, and the gap between the two is widening.
The Ladder describes five levels:
- At Level 0, AI is ungoverned: departments experiment independently, shadow tools proliferate, and the firm accumulates risk without benefit.
- At Level 1, AI automates discrete tasks. Cost is reduced, but no advantage is created; every competitor with the same vendor reaches the same output.
- At Level 2, AI improves the quality of decisions: market intelligence, demand forecasting, cross-border pricing models. A Level 1 firm makes the same decisions faster. A Level 2 firm makes better ones.
- At Level 3, AI enters the operating model itself and the question answered is how the firm operates; removing AI would require redesigning operations, not merely switching tools.
- At Level 4, the question shifts entirely: why is the firm organised this way? Structure, leadership roles, geographic footprint, and talent architecture are determined by AI capability. Firms at Level 4 are not deploying AI into an existing organisation; they are building organisations whose design AI has shaped from the outset.
At Level 4, the question is no longer ‘how do we use AI?’ It is ‘why are we organised this way — and would we design ourselves differently if we started today?’
What Causes Firms to Climb?
The Ladder is explanatory, not merely descriptive. Five drivers determine movement between levels. Data capability is the rate-limiting factor for Level 1 to Level 2: without integrated, clean data across markets, AI cannot generate reliable decision-relevant insight. Leadership mandate drives the move from Level 2 to Level 3: embedding AI in the operating model requires CEO or CFO sponsorship, not a CIO initiative. Competitive pressure is the most reliable accelerant: firms observe rivals gaining measurable advantage and raise internal urgency in response.
Regulatory clarity paradoxically enables Level 3 investment by resolving governance uncertainty while ambiguity functions as a brake. Talent architecture determines whether Level 4 is reachable at all: reorganising around AI requires AI-native capability embedded across business functions, not concentrated in a single technology team.
The Diagnostic Question
Every leadership team should be able to answer one question: is AI on the strategy agenda or the IT agenda? The answer almost always reveals the level the firm actually occupies, not the level it believes it occupies. The firms that will define the competitive landscape in global business over the next decade are not those with the most advanced models. They are the ones that understood, early enough, that AI strategy is organisational strategy, and built accordingly.
