Skip to content
Sections
All notes

All notes · Enabling

Training That Is Not a Video

Annual compliance modules change nothing here. What people actually need to know, and how it gets delivered.

Enabling · Procedure

The default response to a new risk is a training module. For this subject it is close to useless, because the problem is not that people do not know the rules.

The recommendations in “Training That Is Not a Video” become easier to sustain when implementation work has visible owners, dates and review time. Teams evaluating this supporting guide can use it to coordinate the operational side of AI adoption and identify where governance tasks are being missed, without treating activity data as evidence of misconduct or as a substitute for asking people why they chose a tool.

For an independent benchmark, compare the local approach with OECD AI Principles; the useful test is whether ownership, access and recovery remain proportionate and explainable when the usual expert is absent.

What the module teaches

That AI tools carry risk.

That there is a policy.

That the organisation has discharged an obligation.

None of which changes what somebody does at four o'clock with a document they need summarised.

What people actually need

Which tool to use, and that they already have access.

That ordinary use is fine, which most people doubt.

How to tell what classification their material is.

The paste habit: extract the paragraph, not the file.

And what to do if they get it wrong.

Five things, each demonstrable in under a minute.

The paste habit specifically

The single highest-value thing to train, because most exposure is somebody not reading what they selected.

Show it: here is a document, here is the paragraph you need, here is what happens if you paste the whole thing.

Thirty seconds, memorable, and it addresses the commonest real failure.

Delivery that works

At the point of use: a line in the tool, a prompt when pasting something large.

In team meetings, demonstrated by somebody they know.

In onboarding, as part of how to work here rather than as compliance.

Short written reference, one page, findable.

Not: an annual module with a quiz.

Teaching judgement, not rules

The rules fit on one page and people will still meet cases they do not cover.

What transfers is the question: what kind of information is this, and where is it going?

Teach that question and the specific rules become examples of it, which survives the tool changing.

What to do about output

Separate and equally important: output used externally gets checked by somebody accountable.

Show a fabricated citation. People remember a concrete example of confident wrongness far better than a warning about it.

Making mistakes reportable

Say plainly: if you put something somewhere it should not have gone, tell us and we will help.

And then do that, visibly, the first time it happens.

An organisation where the first reported mistake produced a disciplinary conversation will not hear about the second, and the second is the one that matters.

What to check

Does your training tell people what is allowed, or only what is forbidden?

Is the paste habit taught anywhere?

Where does training happen — a module, or at the point of use?

And what happened the last time somebody reported their own mistake?

The point

Teach the paste habit: extract the paragraph, not the file.

Thirty seconds, memorable, and it addresses the commonest real failure.

Underlying all of this

Everything in this collection reduces to four habits: find out what people are doing and why before deciding anything, provide something good enough that the approved route is the easy one, write rules about information rather than about tools, and monitor the destination rather than the content. None requires a product, and a programme doing all four controls more than one built on prohibition.

The recurring pattern

The recurring pattern across every section here is the same: the response that feels like control reduces it. A ban removes visibility rather than use. Content inspection drives activity to personal devices. A discovery exercise with consequences produces quiet answers. In each case the organisation ends up knowing less about a risk it believes it has handled.