36% Mention AI Skills, Only 12% Feature Them. Put Yours Up Top

If you use AI at work and it isn't near the top of your resume, you're in the majority, and that majority is losing interviews. ZipRecruiter's latest New Hires Survey found only 12% of new hires feature AI skills prominently on their resumes, and only 36% mention them at all. Meanwhile, 74% of employers in a separate ZipRecruiter report say AI skills are a strong advantage or a requirement. Most people have the skill. What they're missing is where they put it on the page and what they offer as proof.

Employers want AI skills far more than resumes show them

Employer demand for AI skills is well ahead of how candidates present them: 74% of employers treat AI skills as an advantage or a requirement, but only 36% of new hires mention AI at all. That gap is an opening for anyone willing to close it.

The employer figures come from ZipRecruiter's AI employer report, published July 29, 2026, which surveyed more than 1,000 U.S. employers. Two more details from it matter:

• 13% of employers say AI skills are required for all roles, not just technical ones.
• 50% expect candidates to already be practical or advanced users.

The New Hires Survey shows what that demand is worth. New hires in roles that listed AI proficiency as required got a median of 10 interviews and 3 offers. The typical new hire submitted a median of 16 applications and got 5 interviews and 2 offers. That works out to about 2x the interviews and 1.5x the offers.

One fair caveat. 4 Corner Resources points out that ZipRecruiter is a hiring marketplace, so a ZipRecruiter study concluding AI is good for hiring also flatters ZipRecruiter's own market. It still judges the sample solid. Read the multiples as directional, but don't ignore them.

Two-thirds of people who use AI at work bury it

Of the 36% of new hires who mention AI, only a third feature it prominently. So roughly two in three candidates who have the skill hide it where nobody reads. That's a placement problem, not a skills gap.

Here's why that matters. A screener reading quickly sees the summary and the first bullets, not line 40. I call that zone the top third: the summary, the first two bullets under your current role, and a short skills line, roughly what shows on screen before anyone scrolls. An AI mention on line 40, or a lone tool name at the end of a skills list, technically exists. Functionally, nobody sees it.

Experienced people bury it more. ZipRecruiter found first-time new hires are twice as likely to list AI skills (65% vs 32%). My read on why: a mid-career analyst or marketer sees AI as a tool for doing the job, the same way they'd never list "email." A new grad sees it as a credential. In 2026 the new grad is closer to what employers are screening for.

Refolk's index: under 1% of marketing managers put AI up top

In Refolk's index of professional profiles, 88,119 U.S. profiles match "Marketing Manager." Only 530 put "AI" in their headline, and just 83 carry a "Generative AI" skill tag. On profiles, the AI signal is even rarer than on resumes.

Some caveats. The title match is broad (it includes some founders and CEOs), and skill tags are self-reported. So these numbers show how people present themselves, not how many of them actually use AI. But self-presentation is exactly what a recruiter searches on.

Who uses which signal is the useful part. In Refolk's sample, the marketing managers with a Generative AI skill tag work at Mastercard, Microsoft, Accenture, and Johnson Controls. The ones with "AI" in their headline lean toward small agencies, founders, and fractional CMOs: shops like Social-Kraft and Fix My Flyer, and profiles titled "Fractional CMO Services." Put "AI" in your headline and you look like an agency or a freelancer. In-house people quietly tag it as a skill.

What that means for you:

• In-house candidates: show AI through outcome bullets, not a headline label. "AI Marketing Strategist" in your headline can read as freelancer branding to a corporate recruiter.
• Consultants and fractional operators: a headline label fits the norm, but you still need a proof bullet underneath it.

The size of the pool cuts both ways. A recruiter searching profiles for AI-fluent marketers finds only a few hundred people who say so up top. If your AI work lives on line 40 or nowhere, you aren't in that search. You're in the other pile, sending applications and hearing nothing back while recruiters can't find you.

Where to put AI skills on your resume

Put AI skills in three places, in this order: one clause in your summary, the first or second bullet under your current role, and a short skills line that names specific tools. Only 12% of new hires do this, and a standalone "AI" section at the bottom does almost nothing.

Summary: one clause tied to the work

Your summary should mention AI tools once, attached to a business outcome. It doesn't need an adjective. Compare:

• Weak: "Results-driven marketer passionate about leveraging AI."
• Strong: "Lifecycle marketer who runs segmentation and campaign reporting through GPT-plus-Sheets workflows, with every send reviewed before launch."

The strong version tells the screener which tool you use, what you use it for, and that a person checks the output.

First bullets: the proof

The first bullets under your current role are what a fast reader actually reads. If AI changed how you do your core work, that change belongs in bullet one or two. Don't put it in bullet six, after the meetings you ran.

Skills line: tools, not adjectives

Name the actual products you use, grouped on one line and tagged with what you use them for:

• "AI tools: [product] (workflow automation), [product] (data analysis), [product] (drafting and editing)"

Skip "AI-savvy," "AI-first mindset," and "familiar with generative AI." None of those name anything a screener can match or a hiring manager can test.

Change the AI line for every posting

Don't send one version everywhere. A posting that stresses workflow automation wants your automation bullet first. A posting about analysis wants your analysis work first. A posting that mentions AI governance wants to see how you review and check outputs. Reordering the top third for each application is tedious, and it's the exact work Refolk takes off you. Paste the posting and Refolk rewrites your own resume for it, moving the AI work that posting cares about into the summary and first bullets.

Write AI bullets with tool, task, result, and judgment

A believable AI bullet names four things: the tool, the task, the result, and the human judgment you applied. Drop the judgment and it reads like the AI did the job. Drop the result and it reads like you played with a chatbot.

Judgment is the part people leave out, and employers are leaning on it. In ZipRecruiter's employer report, employers rate critical thinking (65%), judgment and decision making (59%), and creativity (58%) as more important than a year ago. A bullet that shows you checked, edited, or overruled the model answers that directly.

The formula:

1. Tool: the named product, never just "AI."
2. Task: the recurring piece of work it touched.
3. Result: time saved, output volume, error rate, or a business metric.
4. Judgment: what you reviewed, validated, or decided yourself.

AI skills resume examples by role

Fill the brackets with your real tools and numbers. Never estimate numbers you can't defend in an interview.

• Marketer: "Cut campaign reporting from 6 hours to 1 with a GPT-plus-Sheets workflow; spot-checked every figure against the ad platform before it went to leadership."
• Analyst: "Built a [tool] process for monthly variance commentary, cutting first drafts from [X] days to [Y]; rewrote each narrative to flag the [N] drivers the model missed."
• Designer: "Used [image tool] to produce [N] concept directions per brief instead of [M]; picked and refined the final two against brand guidelines."
• Operator: "Automated vendor intake triage with [tool], clearing a [X]-ticket backlog; wrote the escalation rules for edge cases the model routed wrong."

All four are AI proficiency resume lines for non-technical roles. None of them need code, and all of them are specific enough for an interviewer to ask a follow-up.

The hard part is usually finding the real numbers, not the wording. They're scattered across old resumes, chat threads, and performance reviews. Refolk builds the resume from your own history, so the bullets carry your actual figures, not plausible-sounding invented ones.

Skip "prompt engineering" and name the workflow

Hiring managers barely reward the label. In Resume Genius's 2025 survey, only 16% prioritized writing effective AI prompts. 36% prioritized using AI tools, and 31% prioritized integrating AI into workflows. Writing "prompt engineering" signals that you know the term, not that you've used it to get anything done.

ZipRecruiter's employer data points the same direction. The skills employers now weight more heavily:

• Workflow automation: 60%
• Data analysis: 60%
• AI governance: 56%

So translate:

The governance row is underrated. 56% of employers weighting AI governance more heavily means "I wrote the review checklist" is now a credential. Most candidates never think to claim it.

Vague AI claims now get tested, and can cost you the offer

Your AI claims will get checked. 11% of new hires had to demonstrate AI proficiency in an interview or skills assessment, and ZipRecruiter says that isn't limited to tech roles. In roles requiring AI skills, rescinded offers hit 39%.

ZipRecruiter reads that 39% as employer demand for AI talent outpacing their ability to define what they're looking for. Here's how I'd expect that to play out. A hiring team that can't say exactly what "AI proficiency" means will test for it live, and decide afterward whether what it saw matches what it imagined. A vague claim gives them nothing solid to hold onto. A specific bullet ("cut reporting from 6 hours to 1") gives you something to demo in two minutes.

ZipRecruiter labor economist Nicole Bachaud put it plainly: "New roles are emerging across the labor market, but the bar is rising with them."

There's a second trap. An Insight Global survey found 88% of hiring managers say they can tell when a candidate used AI to write their resume or cover letter, and 54% say they care. A generic, machine-smooth line about your AI skills is the worst of both: it claims the skill and shows you used it carelessly. Your AI bullets should sound like you, with your numbers and your edge cases in them.

Before you apply to a role that lists AI proficiency as required, make sure every AI claim in your top third is one you could demo on the spot.

FAQ

Should I list AI skills if I'm not in a technical role?

Yes. In ZipRecruiter's employer report, 13% of employers say AI skills are required for all roles, and live AI proficiency tests aren't limited to tech jobs. Marketers, analysts, designers, and operators should name the tool, the recurring task, the result, and the review step. You don't need code. You need a specific workflow and a number.

Where should AI skills go on a resume?

In the top third: one clause in the summary, the first or second bullet under your current role, and a skills line that names specific products. Only 12% of new hires feature AI prominently, so putting it there makes you stand out rather than blend in.

Is "prompt engineering" worth listing?

Rarely on its own. Only 16% of hiring managers in Resume Genius's survey prioritized prompt writing, against 36% for using AI tools and 31% for integrating AI into workflows. Describe what your prompts automated or produced. That outcome is the skill employers are paying for.

Should I put "AI" in my profile headline?

It depends on how you work. In Refolk's index, only 530 of 88,119 U.S. marketing managers use "AI" in their headline, and that group leans toward agencies, founders, and fractional CMOs. In-house marketers at companies like Microsoft and Mastercard tend to list it as a skill tag. If you're in-house, put the proof in your bullets and the tools in your skills.