Companies bought a lot of AI over the past 12 months. The technology was the easy part. Getting a building full of busy, skeptical, habit-bound people to actually change how they work turned out to be the hard part.

So a new job appeared to close that gap.

We wanted to know what this job actually is. So we pulled 417 US job postings for AI enablement, transformation, and adoption roles. We coded every one across 60 dimensions. We left out the people building AI features, because that’s a different job. We wanted the internal role. The one pointed at a company’s own employees.

Here are some of the findings that surprised me most.

Nobody agrees what to call it. Across 417 postings, 371 distinct titles. The most common exact title shows up just 9 times. AI Enablement Lead. AI Transformation Director. Head of AI Adoption. The titles vary, but the mandate absolutely rhymes (which we’ll get to in a second).

It barely has a home on the org chart. Only 28.5% of postings even name a reporting line. Same job lives under the CIO at one company, inside HR at the next, parked in a strategy office at a third.

Everyone’s trying to hire a unicorn. 87.1% of postings want at least four of seven core competencies. The typical posting is asking for three jobs in one body. A change leader who can move a skeptical org. A hands-on builder who can stand up working AI. An executive translator who can make the whole thing legible to the people holding the budget. Each of those is a career on its own.

It’s mostly change management. Named outright in 73.6% of postings. Present in 93.3% once you count the ones that describe it without the words. Nearly every company hiring for this wants someone who can get other humans to change how they work.

There is a technical bar, but it has nothing to do with data. Python or SQL shows up in 7% of postings. Hands-on building with AI shows up in 39.8%. The market wants a power user who ships, not a research engineer. Wire up an agent. Prototype an automation. Prove a point. That’s a different person from the one who builds AI systems from scratch, and pricing yourself as the second one is a costly mistake.

And the biggest hidden requirement in the whole set: it’s a product management job in disguise. Two-thirds of postings describe product work. Owning intake, prioritizing a backlog, making roadmap tradeoffs, running launches. Only 25.7% ever say “product management.” Strong product people are scrolling right past these jobs without realizing they’re qualified.

One more number that was interesting. The median pay midpoint across US postings with full salary ranges is $179,640. This is a senior market. The median posting that states an experience floor asks for 7 years. Not entry-level.

We put all of it in a report. 417 postings, 60 dimensions, the five archetypes hiding inside one title, what the market pays, and where the jobs live.

We also pulled out two shorter pieces for the two people on either side of this hire. One for jobseekers, on how to find these roles when they have 371 names and how to pitch yourself for them. One for hiring managers, on how to stop writing a unicorn req and actually fill the role.

The job basically tells us that companies bought AI. The value didn’t show up. Now they’re hiring a human to drag it across the line.

This is one snapshot of a market that’s changing fast. But the pattern was hard to miss once we read enough of them.

Get the full report, plus the jobseeker and hiring-manager guides, here: backengine.com/aienablementreport