AI can arrive quietly inside software you already use, but it still needs ownership and oversight.

A manageable AI pilot can become something much bigger, so it is important to map where it may end up before it starts.

AI can organise evidence and support a decision, but it cannot take responsibility for what happens to the person affected. Important decisions need genuine human oversight from somebody with the information, time and authority to challenge or reject the AI’s recommendation.

Small automated mistakes become significant when they are repeated across a business.

When AI output moves into somebody else’s work, the level of responsibility changes with it.

A quick AI draft may stay with one person, but basic rules and proper checking still matter.

Not every use of AI carries the same weight. The matrix helps you see the difference.