Skip to content
— Inside the assessment —

What the assessment actually
measures

The Institute of AI assesses practice, not recall. There is no multiple-choice paper and no tool checklist. There are five things it wants you to account for.

— The five dimensions —

Five questions your evidence
has to answer

Each one is applied at every grade. What changes between Student and Fellow is the scope of practice the answers are drawn from.

  1. 01
    Framing

    Whether you picked the right problem

    The most consequential AI decision is usually made before any model is chosen. The assessment looks at how you scoped the problem, what you ruled out, and whether you were willing to say AI was the wrong tool.

    “What did you decide not to build, and why?”

  2. 02
    Method

    Whether your approach survived contact

    Plans are cheap. The assessment asks what actually happened: the constraints you hit, the approach you changed, and the reasoning that carried you from one to the other.

    “Where did your first approach fail, and what replaced it?”

  3. 03
    Evaluation

    Whether you know if it worked

    A system nobody measured is a system nobody can defend. The assessment looks for evaluation you designed yourself, against outcomes that matter rather than metrics that flatter.

    “How would you know if this had quietly stopped working?”

  4. 04
    Responsibility

    Whether risk was handled or deferred

    Governance, bias, privacy and impact are treated as first-order parts of the practice. The assessment asks what you did about them while the work was live, not what the policy said.

    “Who could this have harmed, and what did you do about it?”

  5. 05
    Judgement

    Whether you can account for the calls you made

    Every grade turns on this. The assessment is looking for a practitioner who can explain their decisions to someone who was not there, clearly enough that the reasoning can be challenged.

    “Talk me through the call you are least comfortable with.”

— Counted and discounted —

What carries weight, and what does not

  • The decisions you made and defended
    The number of tools on your CV
  • Evidence drawn from real AI work
    Course completion certificates
  • How you handled what went wrong
    A portfolio of projects that all went well
  • Practice assessed against a published framework
    Seniority, job title, or years served
— Your turn —

Try the five questions on yourself

The free self-assessment puts the same dimensions to you in a couple of minutes and tells you which grade your practice supports. No account, no cost. Joining the Institute of AI as a Member is free as well.