260+ AI use cases, ranked to the 20 worth doing first

The standard AI-in-the-enterprise list reads the same everywhere: summarize documents, draft status reports, chat with your data. None of it is wrong, and none of it tells you what to do Monday morning inside a live transformation program, because it isn't organized around how transformation work actually breaks down.

This catalogue is organized the other way — around the six AMIGA dimensions a transformation has to manage regardless of which AI tools exist. A use case that speeds up requirements gathering is a Process application. One that catches data quality issues before cutover is a Data application. Filing 260+ use cases against that structure is what turns a list into something a specific team can act on.

Ranking is the part that actually saves time. Value and complexity are independent axes — plenty of high-value use cases are also hard to ship, and plenty of easy wins barely move the needle. The top 20 are the intersection: real return, short runway. That is where a first pilot should go, not wherever the loudest vendor demo pointed.

Print the top 20, run one, and use the full 260+ as the backlog for what comes after. That is a more honest starting point than trying to adopt AI everywhere at once, which is how most of these initiatives stall before they produce anything measurable.