How to Show AI Impact in Your Promotion Packet

There is a new and genuinely awkward problem in promotion writing. A large part of what many people shipped this year was accelerated by AI, and nobody has told them whether saying so helps or hurts.

So most people do one of two things, and both cost them. They either conspicuously avoid mentioning it, producing a packet that quietly overstates the difficulty of the work, or they lead with the tooling and produce a packet that reads like a vendor testimonial.

The framing that works is the one that has always worked. Committees assess judgement, scope, and outcomes. AI changes how fast you got there. It does not change what they are assessing.

Does using AI make your work less impressive?

Only if the tool was the hard part. For nearly all promotion-relevant work, it was not.

A useful test: if you removed the AI, would the work have been impossible, or merely slower? If merely slower, the achievement is intact and you should describe it as an achievement. If impossible, you have something even better, because you did something your organisation could not previously do at all.

The genuine risk is different from the one people worry about. It is not that a committee thinks less of you for using AI. It is that you describe the tool instead of the outcome, and the committee cannot find the achievement in your own sentence.

How do you write it so it lands?

Every sentence on the right could have been achieved without AI, and the committee does not need to know either way to evaluate it. That is the point. The number is the argument. The method is a footnote.

When should you talk about the AI directly?

There is one case where the tool belongs in the foreground: when your contribution was the judgement about how and where to apply it, not the output it produced.

That is a real and increasingly valuable staff-level contribution:

• You evaluated where AI-assisted work was safe and where it was not, and your organisation adopted that boundary.
• You built the review process that let a team adopt these tools without shipping worse code.
• You worked out which parts of a workflow should not be automated, and saved the team from an expensive mistake.
• You changed how a whole team works, and the productivity change is measurable at team level rather than yours.

Notice these are all about other people. That is what makes them promotion evidence rather than personal productivity anecdotes, and it is the same test that has always separated senior work from staff work.

What about the "everyone has AI now" problem?

This is the sharpest version of the question, and it deserves a straight answer. If everyone on your team has the same tools, does shipping more still count?

Volume alone counts for less than it used to. That is genuinely true and it is worth adjusting to. When output is cheap, output stops being the differentiator, and the things that remain scarce become the whole story:

1. Deciding what to build. Nobody has automated knowing which problem is worth solving.
2. Judgement under ambiguity. Being right about the tradeoff when there is no correct answer to look up.
3. Influence on other people. Getting three teams to agree is not a task you can delegate to a tool.
4. Accountability. Owning the outcome when it goes wrong, which no tool will do for you.

If your promotion case leans on having produced a lot, it is more vulnerable than it was two years ago. If it leans on judgement and influence, AI has not touched it.

The one-line rule

Write the sentence you would have written if the tool did not exist. Then check it is still true.

If it is, that is your packet line, and how you got there is your business. If it is not, you have found something worth knowing before a committee does, and you have time to go and do the harder version.

For the underlying mechanics of turning work into numbers, see how to quantify your impact.