Cognitive Governance
Not just governance of artificial intelligence. Governance of the cognitive and decision-making processes in the interaction between people and AI.
System 0
The cognitive layer where intuition (System 1), analytical reasoning (System 2) and artificial intelligence work together. The concept was introduced by researchers at Università Cattolica di Milano in Nature Human Behaviour.
The four postures
Exploratory, generative, critical and operating. From the white paper Cognitive Governance in the AI Era (UCL School of Management, April 2026). DEXISION puts them to work.
From vision to work
The paths and the platforms translate governance into decisions, competencies and processes.
Cognitive Governance in the AI Era
A Framework for Strategic Decision-Making When Machines Shape Thinking. A white paper by UCL School of Management, Centre for Sustainable Business, April 2026.
Cognitive Governance in the AI Era
The white paper introduces the AI Thinking Quadrants: a research-based framework for preserving cognitive autonomy and authorship of reasoning when AI participates in strategic decisions.
Model Risk
AI governance today focuses on the danger of inaccurate output: hallucinations, biased recommendations, faulty predictions. Necessary work. But insufficient.
Cognitive Risk
The deeper threat is the progressive delegation of epistemic authority from human to machine: it happens when leaders stop verifying AI output because it "sounds right". Better prompts will not solve it: it requires understanding the cognitive posture a leader adopts towards AI.
Cognitive Governance is not AI Governance.
AI Governance manages the machine. Cognitive Governance manages the mind that decides with it.
| AI Governance | Cognitive Governance |
|---|---|
| Compliance, risk and system controls (AI Act, model audits, data protection) | Quality of human reasoning when AI participates in the decision |
| Question: is the system safe and compliant? | Question: is the leader still authoring the decision? |
| Owned by legal, risk and IT | Owned by leadership itself |
The AI Thinking Quadrants: four postures.
Four distinct ways leaders engage AI in strategic decisions. The result of 18 months of fieldwork and a diagnostic survey of over 100 European executives.
Exploratory
Sensemaking, model-led. AI helps clarify the decision before solving it: alternative framings, hidden constraints, facts separated from interpretations.
Generative
Analytical, model-led. AI rapidly produces strategic alternatives, scenarios and options. The leader curates and selects.
Critical
Analytical, human-led. The leader authors the reasoning; AI serves as an adversarial sparring partner to stress-test its logic.
Operating
Sensemaking, human-led. AI translates decisions into executable architecture: decision rights, accountabilities, governance routines.
Source: the white paper Cognitive Governance in the AI Era, UCL School of Management, April 2026. The postures are operated by DEXISION and trained in the workshop for the leadership team.
Patterns that should concern any organisation deploying AI at scale.
66%
of executives default to the Operating Posture
3%
default to the Critical Posture, despite its importance in high-stakes decisions
62%
of leaders begin with a model-led information-seeking move
45%
experienced "convincing but wrong" output that could have caused material issues
Source: diagnostic survey of over 100 European executives, white paper Cognitive Governance in the AI Era (UCL School of Management, April 2026).
From personal skill to organisational capability.
Posture Awareness
Recognising which cognitive posture you are adopting and whether it matches the decision context.
Deliberate Switching
Different decision phases call for different postures: the challenge is keeping the capacity to switch, without being captured by a single mode.
Transparent Attribution
For consequential decisions, tracing what was human-generated, what was AI-generated and how human judgment was applied.
Frequently asked questions.
What is Cognitive Governance?
The discipline that governs how leaders and organisations preserve reasoning authority when AI takes part in strategic decisions. Introduced in the white paper Cognitive Governance in the AI Era (UCL School of Management, April 2026), co-authored by Giovanni Giamminola, it addresses Cognitive Risk, the progressive transfer of judgment from human to machine. It is structural, not an ethical add-on.
How is it different from AI Governance?
AI Governance manages the machine: compliance, model risk, data protection, regulatory alignment such as the EU AI Act. Cognitive Governance manages the human side of the same decision: whether the leader is still reasoning independently or has started outsourcing judgment to fluent, convincing AI output. The two disciplines are complementary, not interchangeable.
Who developed Cognitive Governance?
Cognitive Governance and the AI Thinking Quadrants framework originate in the white paper "Cognitive Governance in the AI Era" (UCL School of Management, April 2026), co-authored by Giovanni Giamminola. The framework is based on 18 months of fieldwork and a diagnostic survey of over 100 European executives.
How does it apply inside a company?
It starts with posture awareness: recognising which of the four AI Thinking Quadrants a leader defaults to, and whether it matches the decision at hand. From there, organisations build deliberate switching between postures and transparent attribution of what was human-generated and what was AI-generated. DEXISION® is the tool that makes all of this operational on real decisions.
The Cognitive Impact Assessment (CogIA) is developed with Professor Oreste Pollicino (Bocconi University): the policy brief.
From the frame to the first step
You start with management: a speech or a Discovery with the leadership team.