AI Governance Starter Kit

The short version

Governance fails in two directions: too little (a free-for-all that leaks data) or too much (a 40-page policy nobody reads). This kit is the pragmatic middle. Three artifacts do most of the work: 1. **A one-page acceptable-use policy** — what's green-lit, what's forbidden, what needs a check first. 2. **A simple RACI** — who approves new AI use cases, who owns risk, who's consulted. 3. **A use-case register** — a living list of where AI touches your business, with an owner and a risk note each. Get these three in place and you've covered 80% of the governance that actually matters, without slowing anyone down. The full kit includes copy-paste templates for all three, a starter risk taxonomy, and guidance on when a use case needs a deeper review.