Blog

Disclosure Is an Audit Trail: How MYT Uses AI

A small line at the end of a post

Lately I have noticed a new kind of line appearing at the bottom of blog posts. A sentence or two saying that AI helped write what you just read.

I added one myself this week, at the end of Sovereignty Is a Governance Discipline:

Written with AI assistance as part of the MYT Governance Lab series. The ideas, editorial review and final judgment are my own.

I have never presented this series as something I wrote entirely alone. But writing that line down, plainly, felt different. It made me ask why it mattered to me.

Where things are heading

This is not only a matter of personal taste. Official guidance and regulation are moving in the same direction.

In Canada, the federal Guide on the use of generative artificial intelligence treats transparency as a core principle. It tells public servants to clearly indicate when content has been produced using generative AI, and to inform the public of any significant use of it. The guide was most recently updated in August 2026.

In Europe, the EU AI Act’s transparency rules began applying in August 2026. Text generated by AI and published to inform the public on matters of public interest must be disclosed. There is an exception when a person has substantively reviewed the content and holds editorial responsibility for it.

That exception is worth noticing. As I read it, the rule does not treat AI-assisted writing as suspect. The concern is unowned writing. What matters is whether a person stands behind the work.

I am not a lawyer, and as a Canadian writer much of this does not bind me directly. But the direction is clear, and it matches something I already believe.

Disclosure is provenance, not confession

For most of this year, the MYT Governance Lab has made one argument in different forms. Systems need stewards. And stewardship has to be provable: who did what, with which tools, on whose authority.

I have written about it for Power Platform assets, for delivery oversight, for security posture, and most recently for sovereign cloud. In every case, the answer was the same. Keep the record. Make the trail visible.

It would be strange to argue for audit trails everywhere except in my own work.

So I do not think of an AI disclosure as a confession. I think of it as provenance. It tells the reader how the work was made, and who is accountable for it. A reader who knows the method can judge the work fairly. Making the method visible from the outset gives the reader a clearer basis for trust.

How MYT uses AI

So here is the standard I work to, written down so it can be checked.

Where AI helps

Research: finding and summarizing sources, which I then open, check and link
Structure: outlining, organizing ideas and shaping drafts
Editing: tightening language and improving clarity
Images: title and illustration images, which are generated, not photographs

What stays mine

The ideas, positions and direction of every piece
The lab work itself: what I build, configure and test in my own environment
Editorial review, final edits and the decision to publish
Accountability for anything that turns out to be wrong

What I don’t do

Use client data or confidential information in public lab demonstrations, blog posts, or marketing materials. Public lab artifacts use mock data.
Present AI-assisted work as written entirely on my own
Use AI to claim experience I have not had

The line on every post

Written with AI assistance as part of the MYT Governance Lab series. The ideas, editorial review and final judgment are my own.

It is a short standard on purpose. A policy nobody can remember is a policy nobody follows.

An open invitation

I am not writing this to call anyone out. Many people are still working out how they feel about these tools, and how much to say about them. That is fair.

But for consultants, practitioners and firms, I think a simple standard is worth having. Not a long policy. Just a few honest lines: where AI helps, what stays yours, and the sentence you put on your work.

It costs very little. It protects the reader. And it protects you, because A record you keep as you work is stronger than an explanation reconstructed later.

Governance starts with our own work. This is mine.

______________________________________________________________________________________________________________________