Clarify AI Policy reuse

## Jobs to be done

Based on https://kongstrong.slack.com/archives/C06GB3KM155/p1789021130935529 there are inconsistencies in the docs describing how to use policies to reuse configuration across multiple entities.

In some places we suggest you must create duplicate policies for every time it is referenced by another entity. I suspect this is DP implementation details leaking into the V2 entity model description we have chosen for the docs.

## Definition of done

- Update AI Policy and AI Model entity docs. If possible check that Hal response is now correct (I'm not sure if it can be pointed at non-prod resources)
- Check other entity docs for similar references and fix where encountered, pay particular attention to FAQs

## Information

Waiting for responses in this thread: https://kongstrong.slack.com/archives/C06GB3KM155/p1789118529573009?thread_ts=1789021130.935529&cid=C06GB3KM155

## Size

- S (Small). Example: fixing a broken link, updating wording in an FAQ or paragraph