Best ChatGPT Prompts for Recruiting: 14 Templates for Recruiters (2026)

These are the best ChatGPT prompts for recruiting in 2026: 14 templates written for recruiters and hiring managers and organized around the full funnel: the intake meeting, the job post, sourcing and outreach, screening, structured interviews, take-home assignments, reference checks, and closing or declining with respect. Each prompt gives ChatGPT the recruiter role and your real context, so you get usable drafts that stay focused on skills and fairness.

The pattern that makes recruiting prompts work is the same one that makes hiring work: decide what good looks like before you look at candidates. The intake brief and the scorecard come first for a reason. Every later prompt, from the job post to the debrief, gets better when you paste in the outcomes and competencies you agreed on at the start, and it stops the model from rewarding whoever writes the most polished resume.

Which prompt for which stage

• Before the role opens: run the intake meeting brief with the hiring manager, then build the interview scorecard from it. These two outputs are the reference for everything after.
• Writing the post: the job description writer, fed with the intake brief rather than an old posting.
• Finding people: the boolean and sourcing strategy for search, then the candidate outreach message, one candidate at a time. Outreach that mentions something specific gets replies; outreach generated in bulk reads like it.
• Screening: the screening question set for phone screens, and the resume to role match for deciding what to probe. Neither should be the decision.
• Interviewing: the scorecard during interviews, the take-home brief if the role uses one, and the debrief synthesizer to turn scattered feedback into a recommendation.
• Closing: reference check questions, then the offer email or the rejection email, and the status update prompt for anyone waiting longer than a week.

What to paste in

Recruiting prompts fail in a specific way: with too little context, ChatGPT falls back on the average job post it has seen, which is long, vague, and full of requirements nobody checks. Keep a role block and paste it above any prompt:

Role: [title, team, level, location or remote policy, salary range]
Outcomes in the first 6 months: [three concrete results]
Must-haves: [no more than five, each testable in an interview]
Nice-to-haves: [clearly labelled as optional]
Deal-breakers: [legal requirements only, such as work authorization or a required license]
Process: [stages, who interviews, expected timeline]

Five must-haves is a deliberate ceiling. Long requirement lists discourage qualified candidates from applying, and most of the items on them never get tested anyway.

Fairness, bias, and the rules that apply

Language models learn from the hiring text of the internet, including its biases. In practice that shows up as gendered or age-coded wording in job posts, a preference for familiar school and company names when comparing resumes, and suspicion of employment gaps. Every prompt on this page tells ChatGPT to ignore names, schools, and gaps, and that instruction helps, but it does not make the output neutral. Keep the model on drafting and structuring, and keep decisions about people with people.

There are also rules to know about. New York City's Local Law 144 requires employers that use automated employment decision tools to screen candidates in the city to commission an annual bias audit and notify candidates. The EU AI Act classifies AI systems used for recruitment and candidate selection as high-risk, with obligations on documentation, human oversight, and accuracy. Using ChatGPT to draft a job post is not what these rules target; using it to score or rank candidates can be. If you are moving toward automated screening, talk to your legal team before the pilot, not after.

Data is the other line. Resumes contain personal data, so use a business plan that does not train on your inputs, remove contact details you do not need, and follow your company's retention policy for anything you paste. The same thinking is spreading through AI tooling generally, where privacy is becoming a default design constraint rather than a setting.

How to tell whether it is working

Measure the funnel, not the drafts. Outreach reply rate shows whether personalisation is real. Application-to-screen pass-through shows whether the job post attracts the right people or just more people. Interviewer agreement on the scorecard shows whether the rubric is specific enough for two people to score alike. Time from final interview to offer shows whether the debrief synthesizer is helping a team decide or just producing longer notes.

Once the offer is signed, the ChatGPT prompts for HR pack picks up the rest of the lifecycle: onboarding, policy, performance, and exit. Founders hiring their first few people will find the ChatGPT prompts for startup founders pack useful for the pitch side of a candidate conversation, and the ChatGPT prompts for customer support pack has de-escalation and saying-no prompts that translate directly to candidate communication. When a job description or scorecard works well, turn it into a reusable template with Prompt Builder so every hire follows the same fair, consistent process.

You are a recruiter who writes job posts that convert. Write a job description for the role below. Include: a short hook, what the person will actually do, what success looks like in 6 months, the must-have skills versus nice-to-haves, and an inclusive close. Cut jargon and ban ninja, rockstar, and guru. Role details: [paste].

Write a short outreach message to a passive candidate for [role]. Reference something specific from their background, explain why this role fits them in particular, and keep it under 90 words. Honest and human, no copy-paste flattery. Candidate background: [paste]. Role: [paste].

Create 8 screening questions for [role] that surface real ability and fit, not rehearsed answers. Mix experience, situational, and motivation questions. For each, note what a strong answer sounds like. Role context: [paste].

Build an interview scorecard for [role] with 5 competencies to assess. For each: a one-line definition, one interview question, and a 1 to 4 rating rubric describing what each score looks like. Keep it specific so two interviewers would score alike.

Write a rejection email for a candidate who [reached which stage]. Be warm, clear, and brief, thank them for their time, and avoid false promises. If they were strong, say so honestly. Under 100 words. Context: [paste].

Compare this resume against the role requirements. List where the candidate clearly matches, where they may fall short, and 3 questions I should ask to close the gaps. Be objective and avoid bias based on names, schools, or gaps. Resume: [paste]. Requirements: [paste].

Draft an offer email for [candidate] for the role of [title]. Include the key terms in [brackets to fill], express genuine enthusiasm, and make the next steps obvious. Warm and clear, not legalistic. Context: [paste].

For the role below, give me 3 boolean search strings to find candidates, a list of places these people likely hang out online, and 2 non-obvious sourcing angles. Role: [paste must-have skills and seniority].

Synthesize this interview feedback from multiple interviewers into a clear hire recommendation. Note where they agree, where they disagree, the main risks, and what would resolve the open questions. Be honest if it is a no. Feedback: [paste].

You are a senior recruiter preparing for an intake meeting. From these notes [paste the hiring manager's request and any old job post], draft an intake brief with: the problem this hire solves, three measurable outcomes for the first six months, at most five must-have skills each phrased so it can be tested in an interview, nice-to-haves, the interview stages and who runs each, and five questions I should ask the hiring manager to close gaps or push back on unrealistic requirements.

Design a take-home assignment for [role] that tests [two or three must-have skills] using a task close to the real work. It must take a strong candidate no more than [2 to 3] hours. Include the brief as the candidate will see it, the materials provided, what to submit, a scoring rubric tied to the scorecard competencies, and a note on what we will not judge (formatting, tools chosen) so candidates do not over-invest.

Write 8 reference check questions for a [role] candidate, focused on [two competencies we are least sure about after interviews]. Ask for specific examples and comparisons rather than ratings, include one question about the conditions in which the candidate does their best and worst work, and add a follow-up for vague answers. Keep the opening and close short and respectful of the reference's time.

Write a short update to a candidate in the [stage] stage for [role] whose decision is delayed by [reason, stated honestly at a level we can share]. Say where things stand, give a realistic date for the next update, and invite questions. Under 80 words, no vague we will be in touch soon.

Adapt this job description for [LinkedIn post / job board listing / community forum such as a Slack or Discord]. Keep every fact and the salary range identical, change only the length, format, and opening to suit the channel and how people there read. Job description: [paste].