You didn’t wait for permission. Almost nobody did.
How to find out where you stand, and what to do before someone asks
A letter here, a summary there, a reply you didn’t want to write twice. AI has been part of your working week for months. What you’ve probably never done is check what you’ve already put into it.
This book is that check. It starts with an audit of the last ten things you sent, scored against three questions: whose information was that, what happens if the answer was wrong, and who needed to know you used it. Then it gives you a method for everything after. By the end you will:
- know which of your recent prompts need a second look
- sort what you handle into five classes of information, and usually move it down a class before it leaves your desk
- place any AI tool on a three-rung ladder using six questions, and match it to your risk, not your budget
- run a four-step check on anything that goes out under your name
- settle what you have to declare, and to whom, in your own profession and jurisdiction
- keep the tasks where AI pays and stop the ones where checking costs more than doing
- finish with a one-page Protocol and a review that keeps it current, because most of what’s in it expires
For doctors, lawyers, engineers, accountants and anyone else whose name goes on the work.
Not for you if you want a prompt library, a tool tutorial or a forecast of where AI is heading.
J.D. Ashby is the pen name of a doctor registered with Ahpra and practising in Australia, who writes about using AI inside the regulated professions, where confidentiality, verification and disclosure apply all at once and getting them wrong has professional consequences.
General information, not legal, medical or other professional advice.
