How I use AI

I teach people to work with AI honestly, so it would be strange not to say plainly how I use it myself. This is a standing note — it covers everything published here, so that individual pages don't need their own disclaimers. Last updated: 16 August 2026.

The short version

I use AI every day — to research, to think alongside, to sharpen drafts, and to help build parts of this site. Everything published under my name is written or rewritten by me. I decide what it says, I check it before it goes out, and I am responsible for all of it, including the mistakes.

Where AI is talking to you

The Mirror is a conversation with an AI, not with me. It says so before you begin. Nothing you write there is stored after you leave, it is not used to train any model, and no one — including me — reads it. The Companion runs in your own browser, and what you tell it stays with you.

Where AI helps, and I stay the author

The essays, the programme materials and the notes here are mine. AI helps me research, argue with an idea, and tighten a draft, in the same way an editor or a good colleague would. The thinking, the judgment and the final words are my own — and under European law, that editorial responsibility is precisely what I am accepting by publishing them.

Images and design

I am not a designer, so the visual work here is made with tools, as most design now is. Where an image, illustration or diagram is generated rather than photographed, I say so on the piece itself. I don't present anything synthetic as a real photograph.

What I won't do

I don't publish AI-written text as though I wrote it. I don't use AI on someone without telling them. I don't put client or confidential material into tools that keep or learn from it. And I don't claim a machine is a person, or a person a machine.

Which AI

Several. I work across commercial models and open source and choose the tool for the task rather than pledge loyalty to one company. What stays constant is not the vendor but the rules — set out in full in the standard I work by.

Why this page exists at all

Much of this isn't legally required. Europe's AI Act asks that people be told when they are interacting with an AI system, and that synthetic media be marked — both of which I do. It does not ask me to announce that a draft was tightened with a tool, any more than anyone declares a spellchecker. I am saying it anyway, once, in one place, because transparency you only offer when compelled isn't really transparency.

Questions

If anything here matters to you and isn't clear, write to me at hello@sagemodeai.com and I will answer plainly.