The single biggest source of AI risk is employees putting the wrong data into the wrong tool. This guide ends that — a clear four-tier classification system that lets anyone tell within five seconds whether a given piece of data can go into a given AI tool.

What's inside

Four classifications (Public, Internal, Confidential, Restricted) with explicit examples, AI tool allowances, and the news-test heuristic ('Would this be acceptable if it appeared in a news article?'). Print and post in workspaces.

How to use it

  1. Skim the table of contents first. Guides reward jumping to the section that matches your current question.
  2. Apply one section per week. Trying to operationalize a guide in a sprint usually buries the insight.
  3. Note where the guide's advice doesn't match your context. That's your edge.