Crosswalk
Our AI Fluency composite, read against the 4D framework
A mapping, not a certification. It exists to show where our measured content touches the four competencies — and, more usefully, where it does not.
Anthropic has published an AI Fluency framework built on four competencies: Delegation, deciding whether, when and how to engage AI; Description, describing goals so that AI produces useful output; Discernment, assessing AI output and behaviour accurately; and Diligence, taking responsibility for what is done with AI and how. The framework is Anthropic's and is published by them here.
Our side of the mapping is limited to two things the instrument actually contains: an objectively scored AI-judgement block, and a self-report AI Fluency composite covering practical use of AI and digital tools. We have read the four competencies against that content rather than the other way around, which is why two of the four rows below say that we measure nothing.
| Competency | As described in the framework | What our content touches | What it does not |
|---|---|---|---|
| Delegation | Deciding whether, when and how to engage AI. | Partly touched. The objective AI-judgement block asks about when to trust, check or avoid AI, and is scored right or wrong against a key rather than against other people. | It is a small block of scenario items answered in an assessment, not an observation of the delegation decisions a person makes in their own work. |
| Description | Describing goals so that AI produces useful output. | Not measured. | The instrument contains no items that ask a person to describe a goal to an AI system, and no rating of anything they have written to one. We do not measure this competency at all. |
| Discernment | Assessing AI output and behaviour accurately. | Partly touched. The same objective AI-judgement block includes judgements about when output should be checked rather than accepted. | Those judgements are made about described situations. Nobody's appraisal of a real AI output is observed or scored. |
| Diligence | Taking responsibility for what is done with AI and how. | Not measured. | The three ranked situational-judgement scenarios in the assessment are collected but not scored — no expert key exists for them yet — so they contribute nothing to any result. There is no measured content for this competency. |
Our mapping, and nobody else's
This crosswalk is our own interpretive mapping. It is not an endorsement, a certification, an approval or a partnership of any kind. The AI Fluency framework is Anthropic's work; Anthropic has not reviewed this page, has had no involvement in it, and has not agreed with anything on it. We make no claim of coverage of the framework and no claim of conformance with it. Read every row as our reading of our own content, which we may have got wrong.
See what is actually measured
The free AI Adaptation Profile takes about twelve minutes and returns every score with its confidence band.
Take the free profile