The operating architecture
Intelligence-Centred Enterprise
The frame is the training data.
Definition
An Intelligence-Centred Enterprise organises work so human and machine intelligence can continuously improve decisions across the whole system.
The argument
Sense → Reason → Act → Learn
The organisation detects change, makes meaning, acts with intent and feeds outcomes back into the system.
Not another layer
This is not an AI team beside the enterprise. It connects strategy, work, data, governance and learning.
The operating loop
Intelligence must move.
An organisation learns only when signals travel through meaning, action and feedback—then improve the next cycle.
The model
Sense → Reason → Act → Learn is the metabolic loop; operating design determines whether intelligence can travel through it.
Metaphor / example
A signal is useless if it cannot reach a decision owner before the moment passes—or if the result never changes the next decision.
Leadership implication
Design decision rights, signal flow, governance and feedback as one system.
Common misunderstanding
ICE is not an AI operating model bolted onto the company. It is a way of organising the company around intelligence.
External evidence and interpretation
What the research shows—and what Mark reads into it.
Published finding
The NBER Reporter (2024) describes an AI system that learned from successful conversations by skilled agents and passed those behaviours to newer agents as suggestions.
National Bureau of Economic Research — The Economics of Generative AI (NBER Reporter 2024, No. 1)Mark’s interpretation
An example of intelligence moving through the organisation—Mark’s Sense → Reason → Act → Learn reading, not the authors’.
Published finding
BCG’s 2026 release reports 42% of companies expect AI agents to hold real decision authority by 2030, but only 5% have the controls to do so safely.
Boston Consulting Group — Applied AI Index 2026 (press release, 30 September 2026)Mark’s interpretation
Decision rights and governance need to be designed as one system with the technology—the core ICE argument.
Co-developed
Mark co-developed the Intelligence-Centred Enterprise framework with researchers and practitioners.
Read the long argument
- Learning Velocity: The New Economics of Enterprise ValueWhy the businesses that learn fastest could capture a disproportionate share of the value created by AI.Read →
- The Return on CourageWhy AI Needs a Different Kind of Leadership. Mark Cameron's 2026 reflection on the leadership vacuum AI has exposed, the limits of the efficiency story and the Return on Courage as a hypothesis for leaders who choose to build the future they want.Read →
- The hard part is youIndividuals are getting faster with AI while their organisations are not. The gap sits with how leaders respond to uncertainty: the Control Trap, the Unlearning Valley and why using AI is not the same as thinking with it.Read →
Questions people ask
What is an Intelligence-Centred Enterprise?+
It is an organisation designed so human and machine intelligence continuously improve decisions through a shared Sense, Reason, Act and Learn loop.
Is this another AI operating model?+
No. It is an enterprise architecture connecting strategy, work, data, governance, decision rights and learning rather than adding a separate AI layer.
Sources
National Bureau of Economic Research — The Economics of Generative AI (NBER Reporter 2024, No. 1)Boston Consulting Group — Applied AI Index 2026 (press release, 30 September 2026)Mark Cameron — AI profileThe Intelligence-Centred EnterpriseResearch library →Bring it into the room
Start a speaking enquiry →