AI Won't Fix Your Culture. It Will Give You More Of What You Already Measure.

In brief

AI is very good at the work you can already measure: tasks, numbers, speed.
It can’t tell what good looks like at your company if you’ve never said it. So it fills the gap with generic answers.
Name your culture first. Then AI can help you use it.

Every leadership team I talk to has an AI conversation going right now. Which tools to buy, which jobs to speed up, how much time it will save. For some, it’s which jobs to cut. That decision should come after you redesign the work, not before. Somewhere in that conversation, someone usually says it will help with the people side too. Better job posts. Faster screening. Cleaner reviews. Training on demand.

I understand the hope. If AI can write the job post, sort the resumes and draft the review, it feels like the hard parts of managing people are getting easier.

They aren’t. The tasks are getting faster. The people problems are staying right where they were.

The myth: AI will fix our people problems

The idea is that AI is smart enough to fill in what’s missing. Give it enough data and it will figure out who to hire, what to train and how to rate people.

But AI works from what you give it. And most companies have only ever given it one half of the picture.

Why AI gives you more of the same

AI learns from what you already measure. Most companies measure results: revenue, output, deadlines, goals hit. Very few have written down how the work should get done at their company.

So AI does what it can with what’s there. It gets very good at the half you measure, and it fills the other half with the most common answer it can find. The most common answer is generic. And generic is the exact problem most companies already have with their culture.

Here’s what that looks like across the employee life cycle.

Hiring. Ask AI to write a job post and it will list the same soft skills every other job post lists. Team player. Strong communicator. Self-starter. Those words were already the problem. They’re the job requirement nobody wrote down, and now they’re written down generically, faster.

Screening. Ask AI to find the best candidates and it will look for what your past hires had on paper. It can’t see whether those hires worked out because of how they did the job, because nobody recorded that part. Here are the four places it shows up in hiring.

Training. Ask AI to build a communication course and you’ll get a good general course. It won’t teach how communication works at your company, because it doesn’t know. That’s the same reason most training doesn’t change behavior.

Performance reviews. Ask AI to help write review comments and it will produce polished versions of “great attitude” and “needs to be more of a team player.” The reviews will read better. They’ll measure the same half of the job they always did.

In every case, AI did its job well. It just didn’t have your culture to work from.

The truth: AI scales whatever you’ve defined

AI is an amplifier. Point it at a clear standard and it helps you apply that standard faster and more consistently than any team could on its own. Point it at a gap and it fills the gap with something generic.

That’s why the order matters. A company that has named its culture can use AI well. It can ask for a job post that describes the behaviors its best people show. It can build training around how work actually gets done there. It can help managers write review comments against a standard everyone shares.

A company that hasn’t named its culture will get faster at the tasks and keep the same people problems: the hires who look right and don’t work out, the training that doesn’t stick, the reviews nobody trusts. Attracting, engaging and keeping the right people doesn’t get easier because the paperwork got faster.

Before you hand the people side to AI

Ask one question: could you tell AI what good looks like at your company, in behaviors, not values?

If the answer is yes, AI can help you use it. If the answer is “not really,” that’s the place to start. Name your culture first, then let AI help you build it into how you hire, onboard, develop and promote.

Start here