An acceptable use policy is a written set of rules that tells your staff and volunteers which tools they may use, what they may use them for, and which data they may put into them. For AI tools, it usually names the approved tools and the kinds of data that must never be typed in. It also says that people should check what the tool produces before they use it, when to tell others that AI helped with a piece of work, and how to report a problem. For example, a food bank’s policy might allow a chatbot for drafting newsletters but forbid pasting in client names or addresses.
Why it matters
Without a policy, people pick their own tools and decide for themselves what is safe to share, which is how shadow AI spreads. A good policy is short and specific, and it comes with training so people remember it. It is also a practical first step toward AI governance. If you are starting from nothing, the policy template and the course lesson on writing one can help. Before you approve a new AI tool, ask the vendor what happens to the information your staff enter. Their answer will show you what your policy needs to forbid.
Published Updated