This is one of the most live questions on any roadmap right now, and the most common mistake is to put AI on the plan as a heading. "AI adoption" sits there as a workstream, broad and exciting and impossible to hold to account, and a year later no one can say whether it succeeded, because it was never defined as anything you could measure.

AI earns a place on a roadmap the same way anything else does. It has to be tied to a specific outcome, with a way to tell whether that outcome happened. Not "adopt AI in operations," but "reduce the time to handle this process by this much, by getting this many people genuinely using this tool." The moment you have to express it that way, the question stops being about technology and becomes about adoption, which is where it always belonged.

I treat AI on a roadmap as an adoption curve, not a launch. The risk with AI is rarely that the tool does not work in a demo. It is that it is deployed, announced, and then used by a fraction of the people who were meant to use it, while the benefits case quietly assumes everyone did. So the roadmap milestones I care about are usage and outcome over time, not the go-live date. Go-live is the start of the real work, not the finish.

Sequencing matters too. AI tends to expose the state of the data, the process and the governance underneath it. Putting an ambitious AI outcome on the roadmap before those foundations are honest is how you end up with an impressive pilot that never scales. Sometimes the most valuable AI item on a roadmap is the unglamorous work that has to come first.

There is also a discipline question that belongs on the roadmap itself. Where AI touches real decisions or real reporting, the controls around it are part of the deliverable, not an afterthought. A roadmap that shows the capability but not the governance around it is only telling half the story.

For a board, the test is simple. Ask what specific outcome the AI on this roadmap is meant to produce, and how you will know if it did. If the answer is a theme rather than a number, it is theatre, and theatre has a way of consuming budget while delivering a story.