What should an AI policy cover?
An AI policy should make everyday decisions easier: which tools are approved, what information must stay out of AI systems, when a human check is needed, and who to ask when the answer is not obvious.
Free AI policy starter
Answer a few straightforward questions and receive a practical, editable policy you can review, adapt and share with confidence.
Review
Here is a quick sense-check before we create the document. Your editable policy will reflect these choices and include practical guidance for how AI should be used, checked and managed in your organisation.
We will prepare a Word document you can download, edit and share internally. If email delivery is available, we will also send a copy to the address you provided.
Ready
Need help reviewing it or want to suggest an improvement? Send a message.
AI policy guidance
An AI policy should make everyday decisions easier: which tools are approved, what information must stay out of AI systems, when a human check is needed, and who to ask when the answer is not obvious.
They can be useful, but staff need plain rules for confidential information, personal data, client material, copyright, accuracy checks and records of important decisions.
A good policy usually needs input from leadership, data protection, IT or security, HR and people who will use AI in real work, so the final version is practical rather than theoretical.