Still Accountable · 7 parts
All writingThe Demo Is Not the System
An impressive AI demo and a system you can ship under audit are different projects. The gap between them is the work. First in a series for regulated teams.
When the Model Is Wrong, You're Still Accountable
When an AI system gets it wrong, your team still answers for it. How regulated teams design for accountability that can't be delegated to the model.
The Data You Can't Paste Into a Prompt
AI and data governance for Australian regulated teams: residency, data sovereignty, and the Privacy Act duties that decide where your data can go.
Your AI Is a Brilliant Junior
AI produces senior-looking code and content with junior reliability, and never learns from mentoring. The review burden is the cost regulated teams underprice.
You Can't Audit a Guess
AI explainability for regulated teams: a model that can't explain why it produced an answer is a liability the moment a regulator asks you to justify the call.
Speed Was Never Your Bottleneck
AI in regulated engineering makes code faster, but speed was never the constraint. Correctness, trust, and sign-off were. Don't optimise what wasn't slow.
Where AI Earns Its Place
A decision lens for adopting AI on regulated teams: sort tasks by what being wrong costs, then act differently in each zone. The finale of Still Accountable.