Technology · AI and automation · Component
AI use-case register
Every place AI or automation is used, with its task, owner, provider, level of autonomy and risk classification.
The deliverable
What it is
The register is the starting point in both the NIST AI Risk Management Framework and ISO/IEC 42001, for the plain reason that uses nobody has listed cannot be overseen. It includes informal uses — staff drafting with a public assistant, a team scripting its own workflow — since those are often the majority.
Each entry refers to the system landscape for the underlying system and to Data Core for the data it draws on. The register adds what is specific to AI: what task is delegated, how much autonomy is given, which model it depends on, and how the use is classified.
One level in
What it is made of
Each element is a constituent part of the component. Follow one to see the attributes it carries.
Use-case description
The task being done or supported, the decision it feeds, and what a good outcome looks like.
4 attributes: Task · Good outcome · Owner · Underlying system
LearnAutonomy level
Whether the system suggests, drafts for approval, acts with review afterwards, or acts alone.
3 attributes: Autonomy · Why this level · Agreed by
LearnRisk classification
The internal risk grade and, where relevant, the EU AI Act category the use appears to fall under.
4 attributes: Internal risk grade · AI Act category · Compliance entry · Assessed
LearnProvider and model
Which provider and model version the use depends on, since behaviour can change when either does.
3 attributes: Provider · Model version · Last changed
Learn
Ask teams what they use, not what they have approved. The informal uses are where the register earns its keep.
The other components in ai and automation
AI acceptable-use policy
What staff may and may not do with AI tools, which tools are sanctioned, and what material may be given to them.
LearnHuman oversight design
For each use, where a person reviews, approves or can stop the system, and what they need in order to do so meaningfully.
LearnAI evaluation record
The test set, the measured quality, the thresholds accepted, and how performance is watched once the system is live.
Learn