The 70/30 DNA™ Starter Check
Ten Questions for One AI-Supported Decision or Workflow
AI can support research, comparison, organisation, drafting and execution.
But when AI influences work that affects clients, employees, money, relationships, reputation or future direction, human judgement must remain active.
This short reflection helps you examine one current AI-supported decision or workflow.
The Starter Check is a reflection tool. It is not a clinical, legal, cybersecurity, compliance or technical assessment.
How to Use the Starter Check
1. Select one current decision or recurring workflow where AI or automation is already involved.
2. Answer each question honestly.
3. Score each response:
Yes = 2 | Partial = 1 | No = 0
The result is a conversation starter, not a formal risk rating or professional certification.
1. Human Purpose
Is the human purpose of the workflow clear before AI begins producing output?
Score: ___
2. Named Human Authority
Is there a clearly identified person who reviews the result, exercises judgement and retains final decision authority?
Score: ___
3. AI Role Boundary
Is it clear what AI may research, organise, compare, draft or prepare, and what must remain human?
Score: ___
4. Information Boundary
Is there a clear rule about what information may enter the AI environment and what must remain private, anonymised or excluded?
Score: ___
5. Visibility
Can the responsible human see the important facts, sources, assumptions, uncertainty, missing information and alternatives behind the AI-supported output?
Score: ___
6. Human Judgement
Does the person reviewing the output have enough experience, context and decision criteria to recognise when the AI may be incomplete, irrelevant or confidently wrong?
Score: ___
7. Correction and Refusal
Can the human question, correct, pause, redirect or reject the output without unreasonable pressure or operational penalty?
Score: __
8. Whole-Workflow Value
Does AI improve the complete workflow after preparation, checking, correction, rework and final approval are included?
Score: ___
9. Transferred Work
Has the workflow been examined to determine whether AI is reducing total work or merely transferring more reading, checking and correction to another person?
Score: ___
10. Consequence Ownership and Recovered Capacity
Is the responsible Human Authority holder clearly named? If AI saves time or attention, has that person consciously decided where the recovered capacity should return, including work, relationships, wellbeing or another meaningful human purpose?
Score: ___
Total Score: ___ / 20
Scoring Guidance & Next Steps
15–20: Several Human Authority Conditions Are Visible
The selected workflow has several visible human boundaries.
Do not assume the workflow is permanently safe, complete or suitable for increased AI reliance. Continue examining:
- important assumptions
- exceptions
- privacy conditions
- review burden
- consequence ownership
9–14: One or More Boundaries Need Strengthening
Some parts of Human Authority are visible, but the workflow may rely on unclear assumptions, passive review or insufficient information boundaries.
Choose the lowest-scoring question and clarify that boundary before expanding AI use.
0–8: Pause Before Increasing Reliance
Human purpose, active review or consequence ownership may be unclear.
Before increasing automation or reliance, pause the selected workflow and identify:
- the named human authority holder
- what AI may and may not do
- what information is appropriate
- what must be reviewed
- when the human must stop or refuse
- who owns the consequence
Important note
A strong score does not guarantee accurate AI output, legal compliance, cybersecurity, privacy protection or a good decision.
It indicates only that several Human Authority conditions appear to be visible in the selected workflow.
Your Next Human Authority Action
Which question received the lowest score?
Question number: ___
What is one boundary you will strengthen before using AI again in this workflow?
My next action:
___________________________________________
___________________________________________
Who must remain responsible when the output becomes real?
Named consequence owner:
___________________________________________
The 70/30 DNA™ Boundary
70% Human Wisdom | 30% AI Support | 100% Human Authority
AI can support execution.
Human judgment must remain active.
Authority remains human.
The 70/30 DNA™ Starter Check questions and scoring structures are original authored programme materials created by Vicki Wu. Public use of the underlying principle does not grant permission to reproduce, white-label, or sell this assessment without authorisation.
When One Workflow Needs Closer Examination
The Starter Check does not certify a workflow or determine whether AI-supported assistance is appropriate.
If the Check reveals one recurring workflow with unclear boundaries, repeated correction, transferred review work or uncertain consequence ownership, you may request a Workflow Assessment through AI Kindred Families.
A Workflow Assessment explores whether the workflow appears suitable for further consideration within the current Domain Avatar Sprint™ boundaries.
No submitted workflow or completed Assessment automatically becomes a Sprint. Final suitability and participation decisions remain human and are reviewed individually.
Continue to AI Kindred Families, the practical application and service pathway. Please do not submit confidential, privileged, personal, regulated or sensitive information.