AACF
Data & Information ยท DAT-01

Data strategy supports AI

The organisation's data strategy states what its planned uses of AI need from data and sets out how those needs will be met.

Why it matters

AI ambitions often assume data that does not exist, cannot be joined up or is not allowed to be used. Writing AI's needs into the data strategy brings those gaps into view early, and lets data investment be planned alongside AI investment rather than after it.

Levels
0 AbsentConfirmed absent: the evidence shows nothing is in place. Not known is recorded as insufficient evidence, not 0.
1 InitialNo data strategy, or one that does not mention AI; data needs are discovered project by project.
2 DevelopingThe data strategy mentions AI in general terms, but does not state what data AI needs or how gaps will be closed.
3 DefinedAn approved, owned data strategy sets out what planned AI needs from data and the actions and funding to meet it, backed by an approved data architecture.
4 ManagedProgress against the data actions that AI depends on is measured and reported, and slippage that puts AI plans at risk is escalated.
5 OptimisingThe data strategy and the AI strategy are revised together, using evidence of which data constraints have most limited AI value.
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. DAT-01.1 An approved data strategy exists with an owner and a review date.
  2. DAT-01.2 The strategy sets out what planned uses of AI need from data.
  3. DAT-01.3 The strategy sets out the actions and funding to meet those needs.
  4. DAT-01.4 An approved data architecture covers how data used by AI is taken in, stored, organised and maintained.
  5. DAT-01.5 Shared definitions are in place for the data used by AI.
Assessment questions
  • Does the data strategy say what your planned uses of AI need from data?
  • Which data gaps are holding back AI today, and who is closing them?
  • When the AI roadmap changes, how does the data strategy change with it?
Evidence an assessor would expect
  • The approved data strategy, with the sections covering AI needs
  • A data action plan or roadmap linked to the AI roadmap, with owners and progress
  • The approved data architecture
How to reach Level 3
  1. Gather what the planned uses of AI need from data, drawing on the AI strategy (STR-01) and roadmap (STR-06).
  2. Compare those needs with the data held today and list the gaps.
  3. Set out actions and funding to close the gaps in the data strategy, and get it approved with an owner and review date.
  4. Document a data architecture showing how data used by AI is taken in, stored, organised and maintained, with shared definitions for key terms, and get it approved.

Common pitfalls

  • Writing the data strategy in general terms without naming the AI uses it serves.
  • Agreeing actions with no funding attached.

Starter artefacts

  • Data strategy with AI needs section
  • Data architecture diagram and description
  • Shared data definitions list
Ownership and sequence
Accountable
data owner
Contributors
executive sponsor, ai governance lead, delivery lead
Depends on
STR-01
Roadmap step
3
Lifecycle
strategy, design
Cross-references
ISO/IEC 42001A.7.2 direct
MITRE AI Maturity ModelP5.1 direct; P5.4 partial
UNESCO AI Maturity Framework1.3 partial; 6.1 direct

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