Culture & Change ยท CUL-08
Public and customer trust
The organisation understands how the people it serves view its use of AI and takes that into account in its decisions and how it explains itself.
Why it matters
An organisation's use of AI can be lawful and still lose the confidence of the people it serves. Knowing their views and expectations early lets it avoid uses they would not accept and explain the ones they would.
Levels
| 0 Absent | Confirmed absent: the evidence shows nothing is in place. Not known is recorded as insufficient evidence, not 0. |
|---|---|
| 1 Initial | Public or customer views on the organisation's AI use are not sought; reaction is discovered only through complaints or media coverage. |
| 2 Developing | Views are gathered occasionally, for example on a single project, but are not used consistently in decisions. |
| 3 Defined | The views of the people the organisation serves on its AI use are gathered through a defined approach and shape design decisions and explanations. |
| 4 Managed | Levels of trust are measured over time, and changes in trust or emerging concerns lead to recorded action. |
| 5 Optimising | The people served help shape decisions about where and how AI is used, and the approach to building trust is improved from measured results. |
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).
- CUL-08.1 A defined approach is in place and used to gather the views of the people the organisation serves on its use of AI.
- CUL-08.2 These views are considered in design decisions on uses of AI.
- CUL-08.3 These views are considered in how uses of AI are explained.
Assessment questions
- What do you know about how the people you serve feel about your use of AI?
- How have their views affected a decision about using AI?
- How would you find out if trust in your use of AI was falling?
- Research, surveys or panel findings on public or customer views of AI use
- Decision records showing those views were considered
How to reach Level 3
- Decide how you will gather the views of the people you serve on your AI use, such as surveys, panels, user research or reviews of complaints.
- Gather those views on a regular basis and for significant new uses of AI.
- Add a step in design reviews where those views are considered, and record the result.
- Use what you learn to shape how each use of AI is explained to the people it affects.
Common pitfalls
- Relying on complaints and media coverage as the only source of public views.
- Gathering views and not recording how they affected any decision.
- Asking only people who already support the use of AI.
Starter artefacts
- Public and customer views approach
- Findings summary from research or surveys
- Design review record showing views considered
Ownership and sequence
- Accountable
- executive sponsor
- Contributors
- business owner, ai governance lead
- Depends on
- CUL-02
- Roadmap step
- 5
- Lifecycle
- strategy, design, operate
Cross-references
| Algorithmic Transparency Recording Standard | Tier 1 Summary information partial |
|---|---|
| ISO/IEC 42001 | A.10.4 partial |
| JSP 936 Part 1 | para 95 partial |
| MITRE AI Maturity Model | P6.4 partial |
| NIST AI RMF 1.0 | GOVERN 5.1 direct; MAP 5.2 partial |
| UK Data and AI Ethics Framework | Societal impact 2 partial; Societal impact 4 partial; Trade-offs 1 direct; SAT 1.14 partial; SAT 1.15 partial |
| UNESCO AI Maturity Framework | 5.2 partial |
direct: this control at Level 3 meets the item. partial: it covers part of it.