AACF
Culture & Change ยท CUL-03

Staff and representative engagement

Staff and their representatives, including trade unions where present, are involved early in decisions about AI that affect their work, and their input is considered and answered.

Why it matters

The people who do the work know where AI will help and where it will cause problems. Involving them and their representatives before decisions are made produces better uses of AI and avoids the resistance that follows when change is presented as already decided.

Levels
0 AbsentConfirmed absent: the evidence shows nothing is in place. Not known is recorded as insufficient evidence, not 0.
1 InitialStaff and representatives hear about AI after decisions are made, if at all.
2 DevelopingEngagement happens on some projects or late in the process, often as consultation on a decided plan.
3 DefinedAffected staff and their representatives are involved at design stage of significant uses of AI through a defined route, their input is recorded, and they hear what changed.
4 ManagedThe timing and effect of engagement are tracked, and uses of AI that skipped it are identified and addressed.
5 OptimisingStaff and representatives help shape the AI agenda itself, and the engagement approach is improved from their feedback on how well it works.
Level 3 criteria

Level 3 is met when every applicable criterion is met. Criteria marked with an overlay apply only to organisations with that characteristic (overlays).

  1. CUL-03.1 A defined route is in place for engaging affected staff and their representatives on uses of AI.
  2. CUL-03.2 Affected staff and their representatives are involved at the design stage of each significant use of AI.
  3. CUL-03.3 Their input is recorded.
  4. CUL-03.4 They are told what changed as a result of their input.
Assessment questions
  • At what point are affected staff involved in a proposed use of AI?
  • Where representative bodies exist, how and when are they engaged?
  • Can you show a case where staff input changed a decision about AI?
Evidence an assessor would expect
  • Engagement records or minutes from staff and representative meetings about AI
  • Examples of designs or decisions changed after staff input
How to reach Level 3
  1. Define what counts as a significant use of AI that needs staff engagement, and add engagement to the design stage of the stage gates (POR-04).
  2. Agree with representative bodies, where they exist, how and when they will be involved.
  3. Hold engagement sessions with affected staff and representatives during design and record what they said.
  4. Report back to those involved what changed as a result of their input, and what did not and why.

Common pitfalls

  • Consulting on a plan that has already been decided.
  • Recording input but never telling people what happened to it.
  • Engaging only with managers and not with the staff who do the work.

Starter artefacts

  • Engagement route description
  • Engagement record template
  • You said, we did summary
Ownership and sequence
Accountable
people lead
Contributors
business owner, executive sponsor
Depends on
CUL-01
Roadmap step
4
Lifecycle
strategy, design, deploy
Cross-references
EU AI ActArt 26(7) direct
NIST AI RMF 1.0GOVERN 5.2 partial; MAP 1.2 partial; MAP 5.2 partial; MEASURE 1.3 partial; MEASURE 4.1 partial
UK Data and AI Ethics FrameworkSocietal impact 2 partial; Trade-offs 1 partial; SAT 1.14 partial; SAT 3.8 partial
UNESCO AI Maturity Framework2.4 partial

direct: this control at Level 3 meets the item. partial: it covers part of it.