What hiring managers are actually asking for

We pulled six months of job postings across knowledge-work roles in Canada and the US, looking for the phrase “AI skills” or some near variant. Then we called ten people who’d written them. Two things stand out.

What hiring managers don’t mean

The phrase doesn’t mean “build a model”. Almost no one outside of dedicated AI/ML teams cares whether you can fine-tune a foundation model. Of the ten leads we spoke to, exactly zero were screening for transformer architecture knowledge in non-engineering roles.

It also doesn’t mean “prompt engineering” in the way the LinkedIn course-sellers describe it. Nobody is hiring for a “Prompt Engineer” outside of a few specialist startups. That title is mostly hype residue.

What they do mean

Three things, in roughly this order:

1. You’ve actually used the tools, recently, on real work. Not “I took a course.” More like: “Last week I used [Claude/ChatGPT/Gemini/Copilot] to draft / summarize / clean / analyze X, and here’s what I learned about when it helped and when it didn’t.” That practical fluency is the floor.
2. You can tell when the output is wrong. This is the judgment piece. AI tools confidently produce confidently-wrong content. The question is whether you spot it. Hiring managers want to know you’ll catch the hallucinated citation, the misread context, the subtle factual error.
3. You can teach it to a colleague. The team multiplier. Someone who’s used the tools and can explain to the rest of the team how to use them effectively is suddenly a force multiplier.

The first one filters out most candidates. The second filters out most of the rest. The third is the bonus that lands offers.

Where most people get it wrong

The biggest single mismatch we saw: candidates listing “AI proficient” on resumes the way they list “Microsoft Office proficient”. As a checkbox.

That gets them filtered into a pile that no one prioritizes. The candidates who get pulled out of the pile are the ones who can show — a portfolio item, a workflow they automated, a measurable change in their week.

If you’re updating your resume this month, replace “Proficient with AI tools” with one specific sentence about a workflow you changed and what the result was. That’s the move.

Sector-by-sector specifics

We’re going to do separate primers on what this looks like in HR, accounting, marketing, legal, ops, and customer support. Same shape — what hiring managers are actually asking for, sector by sector. Subscribe and we’ll send you each one as it lands.