expertise analyses interventions approach about glossary FR let's talk
· glossary · strategic intelligence & AI

The concepts that structure AI era decisions

AI creates new power dynamics, new dependencies, new risks. These concepts are what every CEO and executive committee must master to decide correctly, not to understand the technology, but to understand what it changes in power and control dynamics.

France & international · SMEs, mid-market, large companies · CEOs & executive committees

AI Strategic Intelligence

The capacity to read what AI is transforming in power dynamics, dependencies, value capture, responsibility and control. It is not a technical skill but a decision skill.

D-C-A-R

tointelligence reading framework under which every AI decision must be evaluated on four dimensions: Dependency (what does the organisation depend on?), Control (what can it still explain, contest, stop?), Advantage (what hard-to-replicate value is it building?), and Responsibility (who answers for the effects produced by AI?). This framework is the core of the tointelligence approach.

AI Strategic Decision

An AI-related decision that commits the organisation's trajectory over several years by creating dependencies, affecting control, modifying responsibility or influencing competitive advantage.

AI Sovereignty

The capacity to choose, steer and reverse AI dependencies at a cost compatible with one's strategy. Sovereignty is not the absence of dependency but its mastery.

AI Dependency

Relationship created by integrating an AI system into an organisation's usage, processes, data, decisions or architecture, making its removal difficult, costly or risky.

Strategic AI Dependency

Dependency that affects the organisation's capacity to decide, negotiate, function, differentiate or fulfil its obligations.

AI Agent

AI system capable of acting autonomously or semi-autonomously: accessing systems, triggering actions, orchestrating tasks. An AI agent is not a more advanced chatbot: it is a delegation of action. Every delegation of action requires defined rights, limits, supervision and explicit accountability.

Rights of Action (AI)

The scope of actions an AI agent is authorised to trigger: access, writes, sends, validations, escalations. Rights of action must be explicitly defined, limited, traceable and revocable.

Capture

Situation in which a dependency becomes so deep that the organisation can no longer exit without prohibitive cost, operational breakdown or major strategic loss.

AI Control

Capacity to explain, supervise, contest, audit, correct, reproduce or stop an AI system or a decision influenced by AI.

Human Accountability

Real capacity for a human to understand, contest, correct or regain control over an AI-assisted or AI-produced decision. Accountability that is merely declarative — where a human signs without understanding — is not effective accountability. The EU AI Act requires effective human oversight on high-risk systems.

Reversibility

Capacity to exit a dependency, replace a vendor, modify an architecture or resume a process without cost incompatible with the organisation's strategy.

Exit Cost

The total costs — technical, contractual, human, organisational, temporal, financial and strategic — required to exit an AI dependency.

AI Vendor Lock-in

Situation in which an organisation can no longer change its AI vendor without high operational, financial or strategic cost. AI lock-in affects models, data, prompts, workflows, contracts, skills and processes.

Shadow AI

Undeclared, unvalidated or ungoverned AI usage in the organisation. Shadow AI is often the symptom of absent or inadequate governance.

AI Strategic Debt

Accumulation of ungoverned AI tools, contracts, exposed data, dependencies or decisions that become progressively costly to correct. Ungoverned AI POCs today are the AI strategic debt of tomorrow.

AI Governance

System of decisions, responsibilities, rules, escalations, indicators and supervision mechanisms enabling the organisation to control its AI usage.

Board AI Governance

AI governance carried at executive level for decisions that engage strategy, responsibility, dependencies and control.

Build / Buy / Partner

Arbitration by which an organisation chooses to build an AI capability, buy a solution or partner with a third party. The real subject is not the mode of access to the capability but the level of control retained.

Own / Orchestrate / Delegate

A more strategic framework than build/buy/partner. The organisation decides what it must own, what it must orchestrate and what it can delegate.

Strategic Data Asset

Data or dataset capable of creating competitive advantage because it is unique, difficult to replicate, exploitable and close to a critical decision or process.

Defensive Data

Data protected primarily to avoid a risk, without necessarily creating an advantage.

Offensive Data

Data exploited to create differentiating capability, improve a decision or strengthen competitive advantage.

Cognitive Dependency

Situation in which the organisation formally retains human decision-making, but where judgement is progressively pre-oriented by an AI system that decision-makers no longer genuinely contest.

Augmented Decision

Human decision informed by an AI system, with capacity for contestation, supervision and retrieval of control.

Delegated Decision

Decision taken or strongly determined by an AI system. It requires high levels of governance, documentation, supervision and responsibility.

AI Advantage

Capacity to create defensible value through AI, difficult for competitors to replicate. Using a standard AI tool does not in itself constitute an AI advantage.

Captured Value

The share of value created by AI effectively retained by the organisation, as opposed to value captured by vendors, platforms, integrators or competitors.

EU AI Act

European regulatory framework structuring AI obligations according to usage, risk levels and actor roles. For executives, it must be read as a framework of control, responsibility and proof, not only as a compliance constraint.

Critical AI System

AI system whose failure, error, bias, unavailability or opacity can affect an important decision, an essential process, a regulatory obligation, a customer relationship or the organisation's reputation.

Board AI

Decision level where AI subjects cease to be technical and become strategic: dependencies, governance, responsibility, data, vendors, differentiation and control.

Augmented Employee

Employee whose judgement, autonomy, competence and contribution are strengthened by AI. An employee is not augmented because they are given an AI tool. They are augmented if their judgement remains effective and their accountability is preserved.

Responsible Substitution

Explicit organisational decision to replace a human capability with an AI system, accompanied by an analysis of impacts on work, skills and the trajectory of the affected employees. It must be assumed, decided and governed.