AI Governance Canvas
About this resource
Governance discussions expand to fill whatever time they are given, and frequently end with a scheduled follow-up rather than a decision. A canvas constrains the conversation to a single page, which forces the group to choose rather than to explore.
This landscape canvas maps the whole AI governance position in one session: what is in scope, where the boundaries sit, what data may be used, where human oversight is required, who owns what, and when it gets reviewed. It is the working companion to the AI Governance Starter Guide, which explains the eight questions; this is where a group answers them together, in front of each other, in one sitting.
A canvas constrains a discussion that would otherwise expand indefinitely. The space limit is the feature, not a compromise.
How to use it
Print it large, put the right people around it, and complete it in one sitting. The people who need to be there are an executive sponsor, an operations lead, someone from IT or security, and at least one manager who does the governed work. Blank fields at the end of the session are the output: they show precisely where the organization has not decided, which is more useful than a document that papers over the gap.
Inside the download
What is governed and what is deliberately not, in one view.
Acceptable, restricted, and prohibited, side by side.
What may be entered, in your own categories.
Where human judgment is required, mapped against the work.
Common questions
A single-page visual layout covering the main governance decisions, designed to be filled in by a group in one session. The constraint of a single page is what makes it work: it forces a leadership team to choose rather than to keep exploring.
One session of ninety minutes to half a day, provided the right people are present. It takes considerably longer when the group lacks either an executive sponsor who can decide or someone who actually performs the work being governed.
An executive sponsor with authority to decide, an operations lead who understands how the work runs, someone from IT or security for the technical assessment, and at least one frontline manager. Sessions run without that last person produce rules that turn out to be unworkable.
Free. No email list, no follow-up sequence.
More from the resource library, chosen because they pair well with this one.
Eight questions to answer before you write an AI use policy, plus the governance stack and the policy test.
A quarterly test of whether governance exists in practice, written as observable conditions.
Ten questions for directors approving something they cannot evaluate, with what a good answer sounds like and what should concern you.
Tell us a little about yourself and this document is yours. We don't have an email list. We collect this so we know who's finding our resources useful, and that's it.
You're downloading
this document
Your information is kept confidential and never shared with anyone.
