Useful professional knowledge becomes valuable when it changes what you do next. This guide is designed to be read, reflected on and applied.
AI assistance does not remove responsibility
The person using an AI-generated output remains responsible for checking facts, tone, bias and suitability. Treat generated content as a draft that requires professional judgement.
The higher the impact of a decision, the stronger the human review should be. An email draft and a recruitment recommendation do not carry the same level of risk.
Use a simple safety checklist
Before using an AI tool for workplace tasks, understand what information is allowed to leave your organisation.
- Do not paste confidential or personal data into an unapproved tool
- Verify important claims against reliable primary sources
- Document meaningful human review
- Disclose AI assistance when policy or context requires it
Build better prompts from better thinking
Clear objectives, relevant context and explicit constraints produce more useful outputs. Strong prompting is less about clever phrases and more about defining the work accurately.
Ask the tool to identify uncertainty and assumptions. This makes review easier and helps prevent confident but unsupported conclusions.