How to Use AI to Write a Job Description That Actually Attracts Good Candidates

The short answer: Give ChatGPT or Claude the role details and ask for a job description. You’ll have a solid draft in under two minutes. The work on your end is reviewing it, adding what makes your company specific, and adjusting any requirements that don’t match what you actually need.

Most job descriptions are bad. They’re either a recycled version of last year’s posting, a copy-paste from a job board with the company name swapped in, or a wall of bullet points written at 4pm by someone who needed to hire someone yesterday.

AI doesn’t write inspired job descriptions. But it writes competent, well-structured ones quickly — and for most roles, that’s a significant improvement over what was there before, achieved in a fraction of the time.

Here’s how to do it properly.

What to give the AI before you start

The quality of the output depends entirely on what you put in. Before you open ChatGPT or Claude, gather:

Job title — the actual title, not the internal code
Seniority level — junior, mid-level, senior, lead, manager
Key responsibilities — what this person will actually do, in plain terms
Must-have skills or experience — the genuine requirements, not a wishlist
Nice-to-have skills — what would make someone stand out but isn’t essential
Team context — who they’ll work with, who they’ll report to
Company or team tone — formal, casual, startup energy, corporate, etc.
Any specifics — remote/hybrid/in-office, location, salary range if you’re sharing it

You don’t need all of these in polished form. Notes are fine. The AI will turn them into a draft.

The prompt to use

Open ChatGPT or Claude and paste something like this:

“Write a job description for a [job title] role at a [type of company — e.g. fast-growing SaaS startup / mid-size marketing agency / hospital network]. The person will be responsible for [list key responsibilities]. They should have [must-have skills/experience]. Nice to haves include [list]. The team is [describe team context]. The tone should be [professional but approachable / formal / conversational]. Include an About Us section, responsibilities section, requirements section, and a benefits/perks section. Aim for around 400-500 words.”

Within seconds you’ll have a complete draft with appropriate structure, clear language, and all the standard sections.

What to review and personalise

The AI draft will be structurally solid but generically human. Your job is to make it specific:

The About Us section. AI writes placeholder company descriptions. Replace it with something that actually says why someone should want to work there — the specific mission, recent wins, what the team culture is actually like, not what it sounds like in a brochure.

The requirements list. AI tends to pad this. Go through each requirement and ask: is this actually necessary? Unnecessarily long requirements lists discourage qualified people from applying — especially candidates from underrepresented groups who are statistically more likely to self-select out if they don’t meet every listed criterion.

The responsibilities. Make sure they reflect the real job, not an idealised version. If the role involves a lot of coordination work that nobody glamorises in job postings but that’s actually 40% of the job, say so.

The tone. Read it aloud. Does it sound like your company, or like every other job posting? Adjust as needed.

Asking AI to improve what you already have

If you have an existing job description you want to improve rather than starting from scratch, paste it in and ask:

“Rewrite this job description to be clearer, more concise, and more appealing to strong candidates. Remove any jargon or unnecessarily corporate language. Keep all the key requirements but make sure the tone is [professional but approachable].”

Or for inclusion specifically:

“Review this job description for language that might discourage applicants from underrepresented groups. Flag anything that seems unnecessarily restrictive and suggest more inclusive alternatives.”

AI won’t eliminate all bias from a job posting, but it’s a useful pass before it goes live.

Generating interview questions at the same time

Once your job description is done, ask the AI to generate interview questions tailored to it:

“Based on this job description, suggest 10 interview questions — a mix of competency-based and situational questions — that would help assess the key requirements.”

This takes less than a minute and gives you a starting set of questions that actually relate to what you wrote in the posting — instead of generic questions recycled from a template.

A complete workflow in under 15 minutes

Gather your notes on the role (5 minutes)
Paste into ChatGPT or Claude with the prompt above (1 minute, draft appears)
Review and personalise the About Us and requirements (5 minutes)
Ask for inclusive language review if needed (1 minute)
Ask for interview questions while you’re there (1 minute)
Done — post it

The job description that used to take a few back-and-forth email chains and half an afternoon now takes fifteen minutes, most of which is your own thinking rather than the writing.

For more on using AI for work tasks, see how to use AI at work and best AI tools for small business owners.

Frequently asked questions

Can AI write a job description for me? Yes. ChatGPT and Claude can produce a complete, well-structured job description in minutes when you give them the role details. You’ll want to review and personalise it, but the first draft — which is the hardest part — is handled instantly.

How do I use ChatGPT to write a job posting? Tell ChatGPT the job title, key responsibilities, required skills, seniority level, and any specifics about your company or team. Ask it to write a job description in a specific tone (formal, conversational, etc.) and length. Review the output, adjust what doesn’t fit, and you have a posting ready to go.

Are AI-written job descriptions good? The first draft is usually good — well structured, complete, and clear. The parts that need your input are the specifics: what makes your company worth joining, the real culture of the team, and the genuine must-haves versus nice-to-haves for the role. AI gives you the structure; you add the authenticity.

Can AI help make job descriptions more inclusive? Yes. You can ask ChatGPT or Claude to review a job description for language that might discourage applicants from underrepresented groups, rewrite requirements that are unnecessarily narrow, or suggest more inclusive phrasing. AI won’t catch everything, but it’s a useful first pass.