How AI Works & How It's Trained

Before any rules: understand what an AI model is, and what happens to what you type into it.

AI basics~15 min

The big idea

A large language model isn't a database or a search engine — it's a pattern-prediction system trained on vast amounts of text to guess the most likely next words. That training is the key to confidentiality: many tools learn from what users type in, and most retain your conversations on the provider's servers. Understanding this is the foundation for every policy decision in this course — because the risk isn't the model being ‘wrong’, it's where your words go.

Key insight: The data you type into an AI tool may be stored, may be seen by the provider, and — on consumer tiers — may be used to train future models. That is the whole reason the rest of this course exists.

What you'll see

How a model is trained, and the journey your prompt takes once you hit send.

The basics that matter for Defence

Four ideas to carry into the rest of the course. Tap to expand.

Key points to remember

An LLM predicts likely text; it doesn't retrieve facts from a vault.

What you type can be retained, seen, and (on consumer tiers) used for training.

The human using the tool is accountable for the output.

Enterprise/approved tiers exist precisely to control where your data goes.

Back to Dashboard Next: Prompt Engineering