ChatGPT Prompts for HR Managers (65 Free 2026)

A talent acquisition lead at a 420-person SaaS company cut time-to-fill from 41 days to 18 last quarter — not from a bigger team, from a saved library of 14 ChatGPT prompts that pre-screened every inbound resume in under 90 seconds and drafted role-specific interview loops before the requisition hit Greenhouse. Cost-per-hire dropped 34%, offer-accept rate climbed to 87%. Below are 65 prompt patterns HR teams use in 2026 — plus the EEOC and NYC AEDT rules you must respect before pasting a resume into a chat window.

Table of Contents

Contents- What HR Teams Actually Use ChatGPT For in 2026 - EEOC + State Rules Before You Prompt - 20 Resume Screening & Candidate Sourcing Prompts - 15 Interview Prep & Question Prompts - 12 Job Description & Offer Letter Prompts - 10 Onboarding & Internal Comms Prompts - 8 Difficult Conversation Prompts - Best AI Tools for HR & Recruiting 2026 - Honest Limitations - Pro Tips, Warning & FAQs

What

HR Teams Actually Use ChatGPT For in 2026The 2026 SHRM State of AI in HR survey pegs ChatGPT adoption inside US HR functions at 71% — up from 38% in 2024.Talent acquisition leads (89% weekly), followed by internal comms (64%), onboarding (52%), comp reviews (31%). Employee relations and terminations lag at 18% — for good reason.Teams getting real ROI aren't using ChatGPT as a screener that makes decisions. They use it as a drafting engine that produces the first version of a JD, Boolean string, interview scorecard, rejection email, or PIP outline — then a human edits, verifies, and owns the output.That pattern stays inside EEOC and NYC Local Law 144 lines.

E EOC + State Rules Before You PromptThe compliance floor: the EEOC's May 2023 technical assistance on AI in employment selection (binding federal guidance in 2026) says employers are liable for adverse impact even when the tool is a vendor's

If ChatGPT's output disproportionately screens out a protected class, you own it — not OpenAI.NYC Local Law 144 (AEDT) has been enforced since July 2023.If you use ANY automated tool to substantially assist an employment decision for a NYC role, you owe: (1) an independent bias audit within the prior year, (2) a public summary, (3) 10 business days' candidate notice.

Penalties run $500-$1,500 per violation, per candidate, per day.Illinois AI Video Interview Act requires consent + disclosure when AI analyzes recorded interviews. Maryland covers facial recognition.Colorado's AI Act (Feb 2026) extends bias-audit requirements statewide. California's ADS regulations under FEHA (2025) cover any automated decision system in hiring.Guardrails: never paste full resumes with names, addresses, or protected-class markers into a public ChatGPT window — use ChatGPT Team/Enterprise with zero-retention, or redact PII.

Never let ChatGPT make the final yes/no. Log every prompt used in a hiring decision.Run adverse-impact analysis quarterly.

20 Resume Screening & Candidate Sourcing PromptsEvery prompt below assumes you've redacted candidate PII and are using ChatGPT Team or Enterprise

Numbers 1-10 are screening. 11-20 are sourcing.1. Skills gap analysis: "Compare this redacted resume against the JD. Return a 3-column table: required skill, evidence in resume (quote), gap severity (none/minor/major).Do not score the candidate or recommend hire.

Focus only on skill-to-requirement match. Resume: [paste]. JD: [paste]."2. Experience-level flag: "Given this JD requiring 5-7 years of SaaS product marketing experience, extract total years of directly relevant experience from the redacted resume below. List each role, dates, and whether it counts (yes/no/partial) with reasoning. Evaluate role fit, not the person."3.Boolean-string generator: "Build a LinkedIn Recruiter Boolean string for a Senior Backend Engineer, Python + Django + AWS, 6-9 years, US-remote, prior Series B-D SaaS.

Include 3 variations: strict, moderate, wide. Exclude titles containing 'Manager' or 'Director'."4. Resume red-flag scanner: "Review this redacted resume for objective, non-protected concerns only: gaps over 12 months, title inflation, inconsistent dates, mismatched seniority claims.Do NOT comment on names, schools, locations, or demographic signals."5. Skill summary for hiring manager: "Summarize this candidate's technical skills in 4 bullets a non-technical hiring manager can understand.

Focus on demonstrated project outcomes, not job titles."6. ATS keyword extraction: "Extract every hard skill, tool, certification, and methodology from this JD. Return as a comma-separated list for the Greenhouse required-skills field."7.Portfolio review prompt: "Given this candidate's GitHub summary and this Senior Frontend JD, list 5 questions I should ask about their public code to assess depth vs surface familiarity with React 19 and Server Components."8.

Salary-band research: "Using publicly available 2026 data, give me a defensible salary band for a Senior Data Engineer in Austin TX at a 300-person Series C SaaS. Include base, target bonus %, equity range in BPS.Cite Levels.fyi, Payscale, Radford if available."9. Reference-check question set: "Draft 8 reference-check questions for a Director of Customer Success role. Behavioral, open-ended, focused on decision-making under revenue pressure.

Avoid anything that could surface protected-class info."10. Rejection email (post-screen): "Draft a warm, specific rejection email for a candidate who applied to our Senior PM role but has 2 years experience vs 6+ required. 90 words max.Encourage them to apply to associate PM roles when open."11. Passive-candidate outreach: "Write a 4-message LinkedIn InMail sequence for passive Staff Engineers at Stripe, Datadog, and Snowflake.

Each under 90 words. M1: intro + why them. M2: role + comp band. M3: soft follow-up. M4: goodbye with door open."12. Diversity sourcing string: "Build 3 Boolean strings that surface underrepresented cybersecurity talent: (a) HBCU + CS grads 2019-2024, (b) Women in Cybersecurity (WiCyS) members, (c) veteran transition programs.Use org affiliations and program membership only — no race/gender fields."13.

Talent-pool cold email: "Write a re-engagement email to silver-medal candidates from our last round (finalists who didn't get an offer). 3-email drip. Reference their specific role as variable {role}."17.Job fair follow-up: "Draft a same-day follow-up email template for candidates we met at Grace Hopper 2026. 80 words, warm, booth reference variable, clear next step."18. University recruiting outreach: "Draft an outreach sequence to CS program directors at 12 target schools for our 2027 New-Grad SWE cohort.

Include ask for info sessions, resume drops, hackathon sponsorship."19. Boomerang candidate reach-out: "Write a warm re-engagement message to former employees who left in good standing 2-4 years ago and now match our Staff+ IC roles.Reference tenure length as variable."20. Sourcing metrics prompt: "Given last quarter's sourcing data (below), identify which channels produced the highest offer-accept rate and lowest cost-per-hire.

Recommend channel mix for next quarter with reasoning."

15 Interview Prep & Question PromptsGrouped by role type

Every set assumes structured interviewing with defined rubrics — unstructured interviews correlate with 2x adverse-impact risk per SHRM 2025 data.Behavioral (roles: any people manager)21.Manager behavioral: "Generate 6 STAR-format behavioral questions for a First-Line Engineering Manager role. Focus: coaching underperformers, delivering difficult feedback, resolving inter-team conflict.

Include strong vs weak answer signals."22. Accountability without bias: "Draft 5 behavioral questions that assess accountability without penalizing candidates who have caregiving gaps in their resume."23.Disagreement probes: "Give me 4 follow-up probes for 'Tell me about a time you disagreed with your manager.' Surface specifics without leading the candidate."Technical (roles: SWE, data, security)**24.

System design: "Design a 45-minute system design interview for a Senior Backend Engineer. Prompt: URL shortener at 10B/day scale. Include rubric with 4 signal levels and what a Staff-level answer covers that Senior doesn't."25.SQL screening: "Create 8 SQL screening questions from Easy to Hard for an Analytics Engineer. Include expected query, common wrong answers, and what each level of correctness signals."26.

Threat modeling: "Draft 5 threat-modeling questions for a Senior Application Security Engineer. Focus on OWASP Top 10 2025 and supply-chain attacks post-XZ backdoor."Culture-fit (renamed 'values alignment' to reduce bias risk)*27.Values rewrite:* "Rewrite these 3 'culture fit' questions as 'values alignment' questions tied to our published values: Ownership, Customer Empathy, Craft. Job-relevant, not personality-based."28.

Anti-proxy values: "Generate 4 values-alignment questions that avoid proxy-bias risks (no 'beer test', no 'who would you have dinner with'). Grounded in observable work behavior only."Executive29.VP People loop: "Draft an interview loop for a VP of People at a 600-person Series D company. 5 interviewers, 45 min each, distinct themes: talent strategy, comp philosophy, DEI results, HRIS transformation, executive influence.

Include anti-halo cross-check."Non-technical (sales, CS, marketing)30. Sales role-play: "Build a 3-question sales role-play for a Mid-Market AE candidate. Include prospect objection tree, expected discovery moves, and rubric for pipeline vs closing signal."31.Writing exercise: "Draft a 30-minute writing exercise for a Content Marketing Manager candidate — brief, source materials, success criteria, scoring rubric."32.

CSM empathy: "Generate 6 customer-empathy questions for a Senior CSM role that surface how they've handled churn conversations without leading answers."33. Debrief template: "Design a debrief template for a hiring panel. Include structured scoring (1-4 scale, no 3.5s), specific evidence field, and a bias-check prompt for the hiring manager."34.Candidate prep email:** "Write a candidate-prep email 48 hours before their onsite.

Include agenda, interviewer LinkedIn links, format, accommodations line, and what we're evaluating."35. Interviewer training: "Draft an interviewer training memo for a first-timer covering: structured questions only, note-taking mechanics, illegal question list, unconscious bias basics, debrief scoring rubric."

12 Prompts Job Description & Offer Letter Prompts

36.JD debias rewrite: "Rewrite this JD to remove gendered language, corporate-speak, and unnecessary requirements. Flag 'nice-to-haves' that are actually 'must-haves' in disguise. Original JD: [paste]."37. JD from scratch: "Draft a JD for a Senior Product Designer at a Series C fintech. Include: role summary, day-one outcomes, 30-60-90 outcomes, must-haves (max 5), nice-to-haves, comp band, remote policy, EEO statement."38.Internal-to-external JD: "Convert this internal role description into a candidate-facing JD that emphasizes impact over tasks.

Word cap: 400."39. JD adverse-impact audit: "Audit this JD for adverse-impact language (education proxies, arbitrary years-of-experience floors, cultural gatekeeping like 'rockstar' or 'ninja')."40. Channel-specific JD versions: "Generate 3 JD versions for the same role: one for LinkedIn, one for our careers site, one for Hacker News Who's Hiring.Same requirements, different voice."41. Plain-English JD: "Draft a JD in Grade 8 reading level for a Warehouse Team Lead.

Emphasize schedule predictability, benefits, growth path."42. Internal-mobility JD: "Write an internal-mobility JD for a Senior Engineer moving to Engineering Manager. Assume internal candidate. Include what changes, what stays, what support they'll get."43. Standard offer letter: "Generate a standard offer letter for a US-remote Senior Software Engineer.Include base, target bonus, RSU grant with 4-year vest / 1-year cliff, benefits summary, start date, at-will language.

Flag any state-specific requirements for [state]."44. Counter-offer response: "Draft a counter-offer response email for a candidate with a competing offer. Warm, direct, specific to what we can and can't move on. Include escalation path if we need approval."45. Negotiation script: "Write an offer-negotiation script for a recruiter having their first push-back conversation on comp.Include listen-then-respond framing, 3 common objections, and what NOT to say."46.

Signing-bonus clawback: "Draft a signing bonus clawback clause for offers over $50K signing. Include prorated repayment schedule for departures within 12 months."47. Role scorecard: "Generate a Topgrading-style scorecard for a Head of Engineering hire. Mission, 5 outcomes, 8 competencies with weight, cultural anchors."

10 Prompts Onboarding & Internal Comms Prompts

48.New PM 30-60-90: "Build a 30-60-90 day plan for a new Senior Product Manager. Include specific deliverables, stakeholders to meet, and success measures. Mid-size SaaS."49. Week 1 schedule: "Draft a Week 1 onboarding schedule with 12 blocks (welcome, tooling, meet-team, product overview, customer shadow, first quick win). Timezone: US-Pacific."50. Manager onboarding checklist: "Write a manager's-guide-to-onboarding checklist covering pre-Day-1 through Day 90.Include equity, PTO, benefits enrollment triggers, and stay-interview cadence."51.

New-hire announcement: "Draft a company-wide announcement for a new VP of Marketing hire. 200 words, LinkedIn-post-ready, includes their prior company and 30-day priority."52. All-hands recap: "Convert this 45-minute all-hands transcript into a bulleted internal recap for people who missed it. Include decisions, action items, and unresolved questions."53.PTO policy change email: "Draft a change-management email announcing a shift from unlimited PTO to a 25-day accrued policy.

Empathetic, specific rationale, transition timeline, FAQ section."54. Return-from-leave checklist: "Generate a returning-from-parental-leave onboarding checklist for both the returner and their manager."55. RIF comms plan: "Write an internal comms plan for a 15% RIF affecting engineering and marketing.Sequenced messaging, all-hands script, remaining-team message, alumni network invitation."56.

Mid-year check-in: "Draft a mid-year check-in template for managers. Emphasize career growth over performance. 6 open-ended questions."57. Stay interview: "Create a stay-interview script for a manager whose skip-level engineer is a flight risk. 20 minutes, 5 questions, follow-up commitments."

#8 Prompts

Difficult Conversation PromptsThese are the highest-risk prompts in the library.Every output MUST be reviewed by employment counsel or a senior HRBP before delivery. Do not paste real employee names into a public model.58. PIP outline: "Draft a PIP outline for an underperforming Senior AE missing quota 3 consecutive quarters. Include measurable milestones (30/60/90), support resources, and consequences of non-completion.

US at-will state."59. PIP delivery script: "Write a script for the PIP delivery conversation.Empathetic, specific, no leading language. Include how to respond if the employee cries, gets angry, or asks 'am I being fired?'"60. Termination script: "Draft a termination conversation script for a documented performance case. 8 minutes max. Include 'this is a final decision' language, benefits handoff, and dignity-preserving close."61.

Separation summary email: "Write a separation agreement summary email (not the legal doc) explaining severance, benefits continuation, and non-disparagement in plain English.Flag anything that needs employment counsel review before sending."62. Comp-adjustment script: "Draft a comp adjustment conversation script for an employee getting a below-market raise this cycle due to budget constraints.

Honest, no over-promising, retention risk acknowledged."63. Promotion denial: "Write a script for delivering a promotion denial to a Senior IC who expected the Staff bump this cycle.Focus on specific growth areas, not vague 'not ready' language. 12-minute conversation."64. Conflict mediation: "Draft a workplace-conflict mediation opening script for HR sitting between two team members with escalated tension.

Ground rules, format, confidentiality boundaries, what happens after."65. RTW accommodation: "Write a return-to-work conversation guide for an employee returning from medical leave with new accommodation needs.ADA-compliant framing, interactive-process language, documentation requirements."

Best AI Tools for HR & Recruiting 2026| Tool | Cost / user / mo | Best team size | ATS integrations | Bias-audit trail | EEOC-safe defaults |

| Greenhouse AI (native) | Included in plan | Existing Greenhouse customers | Native | Partial, roadmap Q3 2026 | Conservative, opt-in |For a 50-person startup doing 30 hires/year, ChatGPT Team + Greenhouse's native AI is often enough.For a 2,000-person enterprise hiring 400+ across regulated states, Eightfold or HireVue with published bias audits is defensible.

NYC employers using AI to shortlist need an audited platform or an independent audit of their prompt process ($8K-$25K per cycle) — ChatGPT alone isn't sufficient.

Honest Limitations Limitations

Adverse impact is on you. ChatGPT doesn't run four-fifths-rule analysis on your funnel.If prompt-assisted screening surfaces majority candidates disproportionately, EEOC treats it as disparate impact regardless of intent. Run adverse-impact analysis quarterly using ATS reporting or Trusaic.Employment law hallucination is common. ChatGPT misstates state-specific rules — especially final-paycheck timing, non-compete enforceability post-FTC rulings, and PTO payout at termination.

Never send an employment-law answer without a lawyer or verified source review.State compliance drift. Colorado, California, Illinois, New York, and Maryland have overlapping AI-in-employment rules.What's compliant in Texas may violate NYC Local Law 144. Build a state-map matrix into your process.PII exposure is the biggest own-goal. Pasting resumes with names, addresses, phone numbers, or work-auth status into a public ChatGPT window is a GDPR/CCPA issue.

Use ChatGPT Team or Enterprise with retention disabled, or redact first.

Pro Tips, Did You Know, and Warnings

Tips, Did You Know, WarningPro Tip 1 — Prompt-library governance. Build a shared library in Notion or your HRIS (Rippling, HiBob, BambooHR all support this in 2026) with three tiers: green (safe as-is), yellow (edit before use), red (legal review required).Tag every prompt with reviewed jurisdictions.Pro Tip 2 — The 'redact-first' habit. Before pasting a resume, offer letter, or PIP, run a redaction pass: names → [CANDIDATE], schools → [SCHOOL_A], employers → [COMPANY_1], addresses → [LOCATION].

Takes 20 seconds and eliminates the biggest data-exposure risk in AI-assisted HR.Did You Know? The EEOC's first AI settlement — iTutorGroup, August 2023 — was $365,000 for a recruiting tool that auto-rejected women 55+ and men 60+.The tool didn't 'know' it was discriminating; it pattern-matched training data. Every prompt inherits that failure mode unless you design against it.⚠️ Warning — NYC AEDT audit requirement. If your organization has any NYC roles and you're using ChatGPT (or any AI) to substantially assist screening, ranking, or shortlisting, you owe an independent bias audit within the trailing 12 months, a public summary posted on your careers page, and 10 business days' candidate notice.Enforcement is per-violation, per-candidate, per-day.

Audit vendors in 2026 start ~$8,000; full-service compliance runs $18K-$35K.

FAQs

Can ChatGPT screen resumes without bias?

Not on its own. ChatGPT inherits bias from training data and prompt phrasing. It can draft structured comparisons against a JD, but any ranking, scoring, or shortlisting step needs human review, a documented rubric, and quarterly adverse-impact analysis.

Is AI resume screening EEOC-compliant?

AI screening is legal, but you own the outcomes.EEOC's 2023 technical assistance says employers face disparate-impact liability regardless of vendor.

Compliance means four-fifths-rule analysis, decision logs, and a human decision-maker owning every hire/reject.

What's NYC Local Law 144?

NYC's AEDT law, enforced since July 2023. If AI substantially assists employment decisions for NYC roles, you need: (1) an independent bias audit within the past year, (2) a public summary, (3) 10 business days' candidate notice.Penalties: $500-$1,500 per violation, per candidate, per day.

Can I use ChatGPT to write termination letters?

For drafting, yes.

Never paste employee names or facts into a public model. Use ChatGPT Team or Enterprise. Every termination letter needs employment counsel review for state-specific final-paycheck, benefits-continuation, and separation-agreement requirements before delivery.

Which AI tools integrate with Greenhouse or Workday?

All six tools in our table integrate with Greenhouse.Workday: Paradox, Eightfold, HireVue, Textio.

Greenhouse's own AI Suite (Q1 2026) is native. Eightfold and Paradox have the most mature Workday connectors.

How do I run a bias audit on my prompts?

(1) Hire an independent auditor — 2026 rates start ~$8K single-role, ~$25K function-wide (Trusaic, Holistic AI, BABL AI). (2) Use a platform with built-in audits. (3) Internal adverse-impact analysis quarterly — necessary but not sufficient for NYC AEDT.

Can ChatGPT write a Performance Improvement Plan?

As a draft, yes.As the final, no.

PIPs are high-litigation-risk. Use ChatGPT to structure milestones and support resources (prompt 58), then have HRBP and employment counsel review before delivery.

What's the ROI on prompt libraries for HR?

Per 2026 SHRM data, HR teams with governed prompt libraries report 28-45% time savings on drafting work and 15-30% reduction in time-to-fill from faster shortlisting drafts. Cost-per-hire drops as sourcing outreach and JD iteration collapse from days to hours.

Should I disclose ChatGPT use to candidates?

In NYC, yes — Local Law 144 requires it.Illinois, yes for video-interview AI.

Maryland, yes for facial recognition. No explicit federal requirement — but LinkedIn's 2025 Talent Trends survey shows 62% of candidates prefer employers who disclose AI use. Best practice: disclose on your careers page.

How often should I refresh my prompt library?

Quarterly for content. Annually with employment counsel review for legally-sensitive prompts (PIPs, terminations, offer letters, accommodations).Immediately when a relevant state passes new AI-in-hiring legislation.

Ready to

Build Your HR Prompt Library?The 65 prompts above are a starting point. The version that actually cuts your time-to-fill is the one tailored to your ATS, comp bands, states, and voice. Browse the complete HR-tagged library at /prompts?category=hr — every prompt free, tested, copy-paste ready.Related reading on PromptSpace: Full prompt library · ChatGPT prompts for small business owners 2026 · Best AI tools for job seekers & resumes 2026External references (verify current guidance): EEOC technical assistance on AI in employment selection · SHRM AI in HR resource center · HR Dive coverage of AI-in-hiring law · NYC Consumer & Worker Protection AEDT rules