Assess what your AI decisions
create, cede
and commit.

The D-C-A-R grid — the four dimensions

D — Dependency

What does the organisation depend on? At what level? What is its exit capability?

C — Control

What can it still explain, contest, stop, audit, supervise?

A — Advantage

What value is it building that is difficult to replicate? Or does AI mostly benefit the vendor?

R — Responsibility

Who answers for AI-produced effects? Is responsibility effective or merely declarative?

Each dimension is assessed on a mastery scale: high, partial, low or absent. The overall score produces a reading of the organisation's real exposure and action priorities.

What we analyse

This intervention applies to a specific AI decision (before signing), an existing AI systems portfolio (complete diagnostic), or an organisation that wants a permanent D-C-A-R governance framework.

• analysis of current and future dependencies by vendor, model, infrastructure, data and process;
• assessment of real control level — supervision, traceability, exit capability, explainability;
• reading of value captured vs transferred to vendors;
• identification of effective vs declarative responsibilities.

• you want an independent assessment of your real AI exposure;
• you are preparing a structuring AI decision and need a clear dependency baseline;
• your board has no common framework for evaluating AI decisions;
• a deployment, acquisition or major partnership requires a structured exposure reading;
• you want to build a permanent D-C-A-R governance framework for future decisions.