1.3 AI for Defence · Module 1 · AI Basics

Confidentiality & the Classification Boundary

The one rule that matters most — and the redaction habit that keeps you on the right side of it.

The hard line~15 min

The big idea

Here is the rule the whole course turns on: OFFICIAL information can be used with approved generative AI; OFFICIAL: Sensitive and above must never be entered into public AI tools. This is explicit in the DTA staff guidance and the PSPF Advisory. Because models retain (and sometimes train on) what you type, anything sensitive that goes in may be impossible to get back. The practical habit is redaction: strip names, numbers and capability detail before you ever paste.

Key insight: Ask yourself before every prompt: ‘Would I be content if this appeared in tomorrow's newspaper?’ If not, redact it, generalise it, or use your approved enterprise tool.

This is the rule that matters most

The classification boundary is the single most important habit in this course. Everything that follows — approved providers, vendor checks, incident response — exists to support this one line. When unsure of a classification, treat it as higher and ask your security advisor.

What you'll see

The classification boundary, and what redaction looks like in practice.

Staying on the right side of the line

What the boundary means in practice. Tap to expand.

Key points to remember

OFFICIAL can use approved AI; OFFICIAL: Sensitive and above never enters public AI.
Models retain — and may train on — what you type; sensitive data may be unrecoverable.
Redact: roles not names, no identifying numbers, abstract capability.
When in doubt, use your approved enterprise tool, or don't use AI at all.