The Town Hall

AI Governance for Business

Policies, risk and good practice for teams using AI.

Lessons

  1. Find the AI you already use

    Most teams are already using AI, often on free personal accounts nobody approved. Before writing any rules, make a quick list: which tools, who uses them, and what for. No blame, just a map.

  2. Traffic lights for your data

    Sort what your team handles into three colours. Green: public information, fine to use. Amber: internal work, only in tools the business has approved. Red: client personal details, health or financial information, passwords and contracts, never typed into AI. Under Australian privacy law, what happens to a client's details is still your responsibility, even inside someone else's tool.

  3. AI drafts, a person decides

    AI can sound sure and still be wrong. Agree which work a human must check before it goes anywhere: anything sent to a client, anything legal or financial, and any decision about a person. Name who checks it, and keep a note of when AI was used.

  4. Choose your tools on purpose

    Business plans usually let you switch off training on your data, control who has access and delete history. Free accounts often don't. Keep a short list of approved tools with the right settings turned on, and look at it again every six months, because these tools change fast.

Your project

Write a one-page AI use policy for your team: the tools you allow, your traffic-light rules for data, and who checks AI work before it goes out. Share it with your team and ask one person to find a gap in it.

Want help rolling it out? Talk to NoBS IT