AI inventory
Learn how to capture and maintain all AI systems used in your organisation.
The inventory is the source of truth for all AI systems in use. This is where you document external tools, internal applications, ownership, and intended use.
Information
Completeness matters more than perfection. Capture systems even if they are only used occasionally or by a single team.
Key fields
Required fields
- System name
- Provider or vendor
- Short description
- Intended purposes
Additional structure
- Category: internal or external
- Role in the process
- Affected business areas
- Responsible owner
- 1
Create system
Register each AI system as its own inventory entry.
- Open inventory
- Create a new entry
- Use a clear system name
- 2
Fill core fields
Document provider, description, and concrete use cases.
- Select provider
- Add description
- State purposes as concretely as possible
- 3
Add context
Assign the system to business areas, categories, and owners.
- Set internal or external usage
- Add business areas
- Store the responsible owner
- 4
Validate entry
Check whether the entry is sufficient for assessments and reporting later.
- Clean up unclear fields
- Review use cases
- Avoid duplicate systems
Maintain the inventory
Filter
Filter by risk class, status, or missing assessments to see priorities quickly.
Search
Search by name, provider, or business area to find existing systems fast.
Edit
Update changes to usage, ownership, or data flows directly in the record. Changes stay traceable.
Typical examples
ChatGPT
- Provider: OpenAI
- Category: external
- Purposes: draft text, research, summaries
GitHub Copilot
- Provider: GitHub/Microsoft
- Category: internal
- Purposes: code suggestions, review support
Canva AI
- Provider: Canva
- Category: external
- Purposes: visuals, marketing assets
Internal assistant
- Provider: internal
- Category: internal
- Purposes: knowledge lookup, process support
