# The AI decision checklist
Theory Y Technologies Pty Ltd | Practical working material

Use this checklist to frame one business decision, not an organisation-wide promise about AI.

## 1. Frame the workflow
What happens today? Who does the work? What result matters? Which exceptions or hand-offs must be preserved?

## 2. Establish the baseline
What is measured, and what is estimated? Include current quality, time, cost, rework and review effort.

## 3. Compare alternatives
Could clearer processes, deterministic rules or better access to information solve the problem without AI?

## 4. Check information and permissions
Which approved sources are needed? Who authorises access? What data must not enter the system?

## 5. Name the accountable people
Who owns the outcome, operates the system, approves consequential actions and can stop the workflow?

## 6. Specify the evidence
Which representative cases, failure modes and acceptance criteria will determine whether the proposal is credible?

## 7. Plan recovery
What happens when information is wrong, a service is unavailable or the output cannot be trusted?

## 8. Decide the next bounded step
Proceed, investigate, wait or stop. Record the reason, the owner and the trigger for reassessment.

This is general working guidance, not legal advice, certification or a client result.
Permission: internal use with attribution. Commercial republication requires written permission.
Contact: info@theory-y.com.au