How this workflow could help
AI governance foundation
Pre-construction team adopting AI tools without classification, approval gates or disclosure language in client outputs.
01 / At a glance
The opportunity this scenario explores.
See the work this workflow could organise and the decisions your team would retain.
Prepared
procurement questionnaire evidence
Hard-block
Restricted-class data at classification gate
Per call
AI usage audit row written
Current pattern
- AI usage informal - ChatGPT in browser tabs, no policy
- Procurement raises AI questions, response takes weeks
- Restricted-class data leaks into prompts unintentionally
- No audit trail when an output is questioned later
With a structured workflow
- Policy + AUP + AI Schedule + classification rules ship together
- Procurement questionnaire has a prepared evidence starting point
- Classification gate hard-blocks Restricted-class data
- Every AI action logged - reconstructable months later
02 / Inside the workflow
How the work could change.
Situation
A pre-construction team had begun using AI tools to assist tender drafting and clause extraction, but had no classification policy, no approval gates and no disclosure language in client-facing outputs.
Procurement teams at two clients had flagged the lack of an AI Schedule in the SOW - risking a retroactive audit scramble.
Discipline installed
R/A/G data classification rules - Restricted (no AI), Amber (gated AI use), Green (approved AI use). Approval gates per classification baked into the tooling.
AI Schedule template added to the SOW pack with disclosure language for client-facing artefacts. AI usage audit log writes one immutable row per AI call.
What could change
An AI Schedule and usage log would give procurement teams a defined starting point for their questions, subject to their own review.
Monthly classification reviews could identify Restricted-class activity and any gaps in the approval record.
Governance gate reinforced
Reinforces every governance non-negotiable: classification, approval gate, disclosure and audit trail. The Policy Pack is the foundation other packs build on.
In practice
What could be prepared
The working setup.
R/A/G classification rules, approval gates, AI Schedule for SOWs, disclosure templates and AI usage audit log.
Potential change
Procurement teams could review AI-assisted outputs against a recorded decision trail, with each AI call logged for audit.
Explore current workflows03 / Your next step
What could clearer workflows unlock for your business?
Bring us the work you want to improve. We can explore where BuiltAI could help, what information it needs and where your people stay in control.
Discuss Your Starting Point