AI Fluency evidence pack

1 Workforce coverage

What this records. Which registered people have been assessed, and when.

These figures are from Q4 2026, which opened on September 26, 2026 and is still open. They are a position as at October 16, 2026: the 2 people with a result in it, out of 67 registered today. Anyone in the organization can still be assessed in this round, so the figures will change until it closes.

Who has taken part is listed by name and department in the appendix, with the date each was last assessed. No individual’s results appear in this document: every figure is over a group of 5 or more.

Results are shown only for groups of five or more, to protect individual privacy. Fewer than 5 people in this organization have been assessed, so no organization-level result is shown.

2 Governance competency

What this records. Measured Governance scores across the assessed population.

Governance measures whether people keep AI use inside policy, privacy and risk boundaries. Scores are 1 to 4 and are the mean of each person’s most recent result.

3 Too little data to call this round either way

What this records. Measured dimension means across the assessed population, and the training focus: the two dimensions furthest below Applied when any group is under level 3.

Too little data to call this round either way. 2 people have been assessed in this round, and results are reported only for groups of five or more.

4 Coverage and results by department

What this records. Assessment coverage per department, with Governance and Verification means.

A dash means the department’s results are withheld. Results are shown only for groups of five or more, to protect individual privacy. A department can also be withheld when showing it would let a smaller one be worked out from the totals. Withheld departments are included in the organization figures.

5 Results by round

What this records. Each measurement round up to and including this one, oldest first, and the organization mean from each person's latest result within it.

No change since the previous round is stated. Q4 2026 has fewer than 5 responses, so the two are not compared.

A dash means the round is under the privacy floor. Results are shown only for groups of five or more, to protect individual privacy. Kinds: Reconstructed is the round made from results taken before rounds existed; Illustrative is fabricated demonstration data; Test is internal test data.