Method
How the audit works, and what it deliberately does not claim.
THE LAB AI Work Audit™ is a structured review of one real AI-assisted workflow. This page explains the shape of it in enough detail that you can judge whether it is worth $500 to you.
Principles
One workflow, not “your AI”
Auditing AI use in general produces a policy. Auditing one real workflow produces corrections. We ask you to choose the workflow where AI output most often reaches customers or outside systems.
The people who run it, not the people who bought it
The gap between how a workflow is supposed to run and how it runs is where oversight fails. The session includes the person who owns the workflow and, ideally, someone who operates it daily.
Walkthrough, not access
We do not log into your systems and we do not ask for customer records or credentials. The audit is a structured conversation with demonstration, which is enough to make the workflow visible.
Findings you can act on within a month
Every audit ends with three corrections in priority order and one 30-day target. Not a roadmap. Not a transformation program.
Say what it is not
The audit is not legal advice, a compliance certification, a security test, or a guarantee that risk has been removed. The orientation score is experimental and not statistically validated. We would rather under-claim.
The 11-Gate Assessment
Eleven questions asked of one workflow.
Each gate is a place where oversight either exists or is assumed. The session works through them in order, with the people who run the work. The detailed scoring and internal working method are proprietary; the gates themselves are not a secret.
- 01The job the AI has been given, stated plainly
- 02The sources it relies on
- 03What it guesses when information is missing
- 04The decisions and classifications it makes
- 05Where output reaches customers or outside systems
- 06Where a person approves — and whether they really do
- 07What the approver can see
- 08How output is verified before use
- 09Whether a specific output can be reconstructed
- 10How failure is noticed and corrected
- 11Whether the workflow can be stopped, and by whom
The score
The DO YA KNOW WHAT UR LOOKIN AT?™ Score is experimental. We say so every time.
The score orients a conversation about one workflow. It is not statistically validated, it is not a benchmark against other businesses, and it is not a certification. If a vendor offers you a validated AI oversight score today, ask them how it was validated. The free Oversight Check uses a simpler version of the same idea, with the same caveat.
Sequence
- 01
Apply
A short application about one workflow. No payment yet.
- 02
Fit review
We confirm the workflow is a fit and tell you plainly if it is not.
- 03
Pay $500
Founding Pilot fee, one business, one workflow.
- 04
Intake
A brief written intake so the session starts warm.
- 05
Audit
60–90 minutes with the people who run the work.
- 06
Findings in 48 hours
Map, gates, checklist, three corrections, one target.
- 07
Readout
A short review of the findings and the 30-day target.
- 08
Your move
You act on it yourself. There is no obligation to buy anything else.
Founding Pilot · $500
Find out what you're looking at.
One AI-assisted workflow, a 60–90 minute walkthrough, and concise written findings within 48 hours: workflow map, approval gates, verification checklist, three priority corrections, one 30-day target.