Where do AI skills go on a resume?

Two to five tools you actually use in a labelled group in the skills section, and one or two experience bullets that prove the important ones. Put AI in the summary only if the role is about AI.

Is prompt engineering a skill to list on a resume?

Not on its own. Show what your prompts produced and how you checked it. A named workflow with a result is a skill; the phrase with nothing behind it reads as padding.

How to put AI skills on your resume

A tool name is noise; a tool attached to a task and a result is a skill. Where AI goes on the page, using AI tools versus building with AI, the bullet that works by role, and what recruiters distrust.

Last reviewed 19 Sept 2026

The short answer

A tool name on its own is noise; a tool attached to a task and a result is a skill. List two to five AI tools you actually use for work in your skills section, then prove the important ones in a bullet: what you used it for, what you checked, and what changed. Skip “AI enthusiast”, “prompt engineering” with nothing behind it, and any tool you could not demonstrate in an interview.

Nearly every posting now says something about AI, and nearly every resume has answered by adding “ChatGPT” to the skills list. Which means the word has stopped telling a recruiter anything. It is the new “Microsoft Office”: assumed, unverifiable, and faintly worrying when it is the most prominent thing on the page.

What still works is the same thing that works for every other skill. Name it where you used it, and say what happened.

From noise to evidence

Noise
Familiar with AI tools such as ChatGPT.
True of everyone who has opened a browser this year.
A claim
AI tools: ChatGPT, Claude, GitHub Copilot
Fine in a skills list. Still says nothing about what you can do with them.
A use
Used ChatGPT to draft first versions of product descriptions for the catalogue.
Now there is a task. But no sign you were any good at it.
Evidence
Set up an AI-drafted, human-edited workflow for 1,200 product descriptions; cleared a 3-month backlog in 5 weeks, with every page fact-checked against the spec sheet.
Task, scale, result — and the judgment you applied, which is the part a hiring manager is actually worried about.

The top rung names the check you did. That line is what separates someone who uses AI well from someone who pastes its output.

Where it goes on the page

Skills section: a short labelled group, only for tools you use regularly for work. Two to five of them.
Experience bullets: the one or two places AI genuinely changed how a piece of your work got done. This is where the claim is proven.
Summary: only if the role is about AI. For most jobs it is a tool you use, not who you are, and leading with it reads as reaching.

Skills · a marketing role

Channels: SEO, paid social, email lifecycle

Analytics: GA4, Looker Studio, HubSpot

AI tools: ChatGPT, Claude, Midjourney

Languages: English, Hindi, Tamil

One group among four, not the headline. Every tool here should be one you can open and use in front of an interviewer.

Skills · a software engineer

Languages: TypeScript, Python, SQL

AI / LLM: OpenAI and Anthropic APIs, retrieval (pgvector), evaluation harnesses, GitHub Copilot

Infrastructure: AWS, Docker, Postgres

Using AI tools, versus building with AI

These are two different claims and a technical reader separates them instantly. Know which one you are making.

Using AI tools — ChatGPT, Copilot, Claude, Gemini, Midjourney, Notion AI — to do your own job faster or better. Relevant to almost every role now. Proven by what you produced.
Building with AI — calling model APIs, retrieval, fine-tuning, evaluation, agents — as part of a product. An engineering skill. Proven by what shipped, who used it, and how you knew it worked.

✕ Blurs the two

Experienced in AI/ML and prompt engineering; leveraged LLMs to drive innovation.

✓ Says which, and proves it

Built a support-ticket classifier on the OpenAI API with a 400-ticket labelled test set; routed 70% of tickets automatically at 93% agreement with human triage.

The second names the model, the method, the test and the result. The test set is the line an engineering manager looks for first.

By role, the bullet that works

✕ Recruiter / HR

Used AI for recruiting tasks.

✓ Recruiter / HR

Drafted role-specific interview scorecards with Claude and reviewed each with the hiring manager; cut scorecard prep from about a day to 2 hours per role.

✕ Sales

Leveraged ChatGPT for outreach.

✓ Sales

Used ChatGPT to research and draft first-touch emails for 30 accounts a week, personalised by hand; reply rate went from about 4% to 9% over a quarter.

✕ Operations / analyst

Proficient in AI-powered data analysis.

✓ Operations / analyst

Wrote SQL with Copilot's help to automate the weekly inventory report, verifying totals against the ERP; retired 6 hours of manual spreadsheet work a week.

✕ Teacher

Integrated AI into the classroom.

✓ Teacher

Built differentiated reading worksheets at 3 levels with ChatGPT, checked each for accuracy; used across 4 Grade 7 sections.

What a recruiter distrusts

“AI enthusiast”, “AI-first mindset”, “passionate about AI” — adjectives, not skills.
“Prompt engineering” as a standalone skill with no output anywhere on the page.
Twelve AI tools in a row. Nobody uses twelve; it reads as a list copied from a listicle.
A two-hour online course certificate as the headline credential.
Anything that implies the AI did the work and you forwarded it. Say what you checked.
A tool you have not opened in six months. It is the first thing they will ask you to show.

Be ready for the follow-up question

An AI line on a resume now reliably gets one question in the interview: show me how you actually use it. Have one real example ready — the prompt or workflow, what came back, what you changed, and what you caught it getting wrong. The last part is the answer. Anyone can get output from a model; the skill a hiring manager is paying for is knowing when it is wrong.

For the rest of the skills section, see how to list skills on a resume. If your question is the other one — whether a recruiter can tell the resume itself was written with AI — that is a different post.

Common questions

Should I put ChatGPT on my resume?: Only if you use it for real work, and ideally in a bullet that says what you used it for, what you checked and what changed. On its own in a skills list it no longer tells a recruiter anything.
Where do AI skills go on a resume?: Two to five tools you actually use in a labelled group in the skills section, and one or two experience bullets that prove the important ones. Put AI in the summary only if the role is about AI.
Is prompt engineering a skill to list on a resume?: Not on its own. Show what your prompts produced and how you checked it. A named workflow with a result is a skill; the phrase with nothing behind it reads as padding.