268. 50 AI Prompts for HR and Recruitment Teams: ChatGPT, Claude & Copilot (2026)

Last Updated: October 2, 2026

AI prompts for HR teams are now the fastest way to reclaim hours lost to repetitive writing tasks. SHRM’s 2026 State of AI in HR report, which surveyed 1,908 HR professionals, found that 39% of organizations have already implemented AI for one or more HR activities — and recruiting ranks as the single most common use case at 27% of companies. Yet 88% of HR leaders report they haven’t seen significant business value from AI tools yet (Gartner, 2026). The gap is not the tools — it is the prompts.

Culture Amp’s 2026 AI in HR study of 264 HR professionals found that 86% already use AI for content creation and 83% for brainstorming — but most are using weak, vague prompts that produce generic, unusable output. A strong prompt specifies the role, company context, format, tone, and word count. That is what every prompt in this library does. Each one is structured, role-specific, and tested across ChatGPT (GPT-5), Claude (Sonnet 4.5), and Microsoft 365 Copilot — the three tools most HR teams already have access to.

This guide covers 50 prompts across seven core HR functions: recruitment and sourcing, candidate screening, interview preparation, onboarding, performance management, HR policy drafting, and employee engagement. Every prompt is copy-paste ready. Every prompt includes a “Use this when” context line and replacement instructions for bracketed fields. And a dedicated data safety section tells you exactly what to never put into any AI tool. Whether you are an HR generalist managing every people process solo or a specialist recruiter closing 20 roles a quarter, this library will save you a full workday every week.

🧠 1. How to Use These Prompts Effectively

Every prompt in this library follows the same structure: role context, task instruction, format requirement, tone guidance, and word count constraint. That structure is what separates a prompt that returns a usable first draft from one that returns a generic paragraph you will delete immediately. Before you copy any prompt, read the “Use this when” line — it tells you the exact situation the prompt was designed for, so you are not force-fitting a performance review prompt into a sourcing task.

Replace every bracketed field with your specific details before submitting. The more specific your replacements, the better your output. “[Job title]” should become “Senior Data Engineer.” “[Company name]” should become your actual company name. “[Department]” should become “People Operations” or “Engineering” — not a generic placeholder. If you are using Microsoft 365 Copilot inside Word or Outlook, you can reference existing documents directly in your prompt — that is where Copilot has a structural advantage over standalone ChatGPT or Claude. For nuanced, compliance-sensitive documents like termination letters or investigation reports, Claude Sonnet 4.5 or GPT-5 are recommended over Copilot for their handling of longer, more careful drafts.

The 2026 HR Prompting Reality: Recruiters who use structured AI prompts save an average of one full workday per week, according to LinkedIn’s 2025 Future of Recruiting report. The output quality difference between a weak prompt and a strong prompt is not marginal — it is the difference between a draft you publish and one you rewrite from scratch.

📋 2. Recruitment and Job Description Prompts

Job description writing is the most common AI use case in HR at 66% of recruiters who use AI (SHRM, 2025). It is also where weak prompts cause the most damage — generic job descriptions attract generic applicants. The prompts below produce structured, inclusive, role-specific postings that reflect your actual company rather than a template anyone could have written. Use them in ChatGPT, Claude, or Copilot interchangeably for this section.

Prompt 1 — Inclusive Job Description from Scratch

“Write a job description for a [job title] role at [company name], a [company size and industry] company based in [location]. Key responsibilities: [list 4–5 duties]. Required skills: [list 4–5 skills]. Nice-to-have: [list 2–3]. Use a professional, inclusive tone. Remove gendered language and unnecessary degree requirements. Structure in sections: Overview, Responsibilities, Requirements, Nice-to-Have, What We Offer. Keep it under 450 words. Suggest two alternative headline options at the end.”

Use this when: Writing a new role from scratch with no existing template.
Replace: [job title], [company name], [company size and industry], [location], all skill and duty lists.
Works in: ChatGPT, Claude, Copilot

Prompt 2 — Rewrite an Existing Job Description

“Rewrite this job description for a [job title] role. Keep it under 400 words. Lead with impact — what will this person actually change or build? Remove gendered or age-coded language, corporate buzzwords like ‘rockstar’ or ‘ninja,’ and any requirements that are not genuinely necessary for the role. Add a short ‘What success looks like in 90 days’ section at the end. [Paste existing job description below.]”

Use this when: Refreshing a stale job posting or improving a low-application-rate role.
Replace: [job title], paste your existing description at the bottom.
Works in: ChatGPT, Claude, Copilot

Prompt 3 — Boolean Search String for Sourcing

“Generate a Boolean search string for finding [job title] candidates who have experience with [key skills or tools]. Optimize for LinkedIn Recruiter search. Include variations for common title formats. The candidate should have worked at companies with [company size or type — e.g., fewer than 500 employees / Series B startups / enterprise FTSE 500].”

Use this when: Starting a new sourcing campaign and needing LinkedIn search strings.
Replace: [job title], [key skills or tools], [company size or type].
Works in: ChatGPT, Claude (best for variations), Copilot

Prompt 4 — Candidate Outreach Message (Cold)

“Write a cold outreach message to a passive candidate for a [job title] role at [company name]. The message should be under 100 words, reference something specific about their background (leave a [PERSONALIZATION PLACEHOLDER] for me to fill in), explain why this role is worth their time, and end with a low-friction call to action. Avoid generic phrases like ‘I came across your profile.’ Tone: direct, respectful, human.”

Use this when: Reaching out to passive candidates on LinkedIn or by email.
Replace: [job title], [company name]. Fill [PERSONALIZATION PLACEHOLDER] manually before sending.
Works in: ChatGPT, Claude, Copilot (in Outlook)

Prompt 5 — Follow-Up Message After No Response

“Write a short follow-up message (under 60 words) for a [job title] role at [company name]. I sent the first message [X days] ago and got no reply. Do not apologize for following up. Reference the original opportunity briefly. Give them one new reason to respond — a recent company win, a team update, or a deadline. End with a simple yes/no question.”

Use this when: Following up on an unreplied sourcing message after 5–7 days.
Replace: [job title], [company name], [X days].
Works in: ChatGPT, Claude, Copilot (in Outlook)

Prompt 6 — Internal Job Posting

“Write an internal job posting for a [job title] opportunity in [department] at [company name]. Acknowledge that we are posting internally first. Highlight growth opportunity and what makes this role exciting. Include application process and deadline: [date]. Tone: warm, transparent, encouraging. Under 300 words.”

Use this when: Posting a role internally before going external.
Replace: [job title], [department], [company name], [date].
Works in: ChatGPT, Claude, Copilot

Prompt 7 — Job Ad for Social Media

“Write three versions of a LinkedIn job ad for a [job title] role at [company name]. Version 1: lead with the mission. Version 2: lead with the compensation or benefits. Version 3: lead with the team or culture. Each version must be under 150 words and end with a clear call to action. Avoid hashtag spam.”

Use this when: Running a paid or organic social campaign to attract applicants.
Replace: [job title], [company name].
Works in: ChatGPT, Claude

🔍 3. Candidate Screening Prompts

Resume screening ranks as the second-highest AI recruiting use case at 44% adoption (SHRM, 2025). The prompts below help you distill applicant information quickly and consistently — without introducing bias by pasting personal identifying information into public AI tools. Always remove candidate names, contact details, photos, and dates of birth before using any AI tool for screening. See the Data Safety section before proceeding.

Prompt 8 — Resume Summary for Screening

“Summarize this resume in 4 sentences. Focus on: (1) total years and type of relevant experience, (2) specific skills that match our requirements for a [job title] role, (3) any clear skills gaps, and (4) career trajectory — is this person moving toward or away from this type of role? Do not infer or speculate. Only summarize what is explicitly stated. [Paste resume text — remove name and contact info first.]”

Use this when: Reviewing a high volume of applications and needing fast first-pass summaries.
Replace: [job title]. Remove all PII from the resume before pasting.
Works in: ChatGPT, Claude (strongest for longer documents), Copilot

Prompt 9 — Compare Multiple Candidates

“Compare these [number] candidates for a [job title] role. For each candidate, assess fit on: (1) relevant experience, (2) required technical skills: [list skills], (3) seniority level match. Output a comparison table with a brief recommendation for which to prioritize for a first interview. Do not make any assumptions beyond what is stated in the profiles. [Paste anonymized candidate summaries below.]”

Use this when: Shortlisting from a pool of screened applicants.
Replace: [number], [job title], [list skills]. Anonymize all profiles first.
Works in: ChatGPT, Claude, Copilot

Prompt 10 — Screening Call Questions

“Write 8 phone screen questions for a [job title] role. Mix: 3 role-specific technical questions, 3 motivation and fit questions, and 2 logistics questions (availability, location, salary expectations). For each question, tell me what a strong answer sounds like in 1–2 sentences. Keep each question under 20 words.”

Use this when: Preparing for an initial screening call with a shortlisted candidate.
Replace: [job title].
Works in: ChatGPT, Claude, Copilot

Prompt 11 — Rejection Email (Post-Application)

“Write a respectful, warm rejection email for a candidate who applied for a [job title] role at [company name] but will not be moving forward. Do not give specific feedback at this stage. Keep it under 80 words. Thank them for their time. Leave the door open for future opportunities. Do not use the phrase ‘We regret to inform you.’ Tone: human, not corporate.”

Use this when: Sending post-application rejections at the screening stage.
Replace: [job title], [company name].
Works in: ChatGPT, Claude, Copilot (in Outlook)

Prompt 12 — Rejection Email (Post-Interview)

“Write a respectful post-interview rejection email for a candidate who reached [interview stage — e.g., final round] for a [job title] role at [company name] but will not be receiving an offer. Acknowledge that they invested real time. Provide one specific, genuine strength you observed (leave a [STRENGTH PLACEHOLDER] for me to fill in). Keep it under 120 words. Tone: direct, warm, honest.”

Use this when: Declining a candidate after one or more interview rounds.
Replace: [interview stage], [job title], [company name]. Fill [STRENGTH PLACEHOLDER] manually.
Works in: ChatGPT, Claude, Copilot

🎯 4. Interview Preparation Prompts

Structured interviews consistently outperform unstructured ones in predicting job performance — yet most hiring managers still wing it. These prompts help you build competency-based question banks, scoring rubrics, and debrief frameworks in minutes rather than hours. For all interview prep prompts, run the output through your own review before use. AI tools do not know your company’s specific competency frameworks unless you tell them explicitly.

For AI tool comparisons relevant to your hiring tech stack, see our Claude vs ChatGPT vs Gemini business comparison to understand which model fits which HR workflow best.

Prompt 13 — Behavioral Interview Question Bank

“Build a structured interview question bank for a [job title] role at [seniority level — e.g., senior, mid-level]. Include 12 behavioral questions covering these competencies: [list 4–6 competencies, e.g., leadership, communication, problem-solving, adaptability]. For each question, include: (1) the competency it targets, (2) what a strong answer looks like, and (3) a probe follow-up question.”

Use this when: Preparing a structured interview for a new or recurring role.
Replace: [job title], [seniority level], [list competencies].
Works in: ChatGPT, Claude (best for rubric depth), Copilot

Prompt 14 — Technical Interview Scorecard

“Build a structured interview scorecard for a [job title] role. Include 5 competency areas, 2 questions per competency, and a 1-to-4 rating rubric for each with behavioral anchors for scores 1, 2, 3, and 4. Format as a table. Competencies to include: [list 5 competencies].”

Use this when: Standardizing scoring across multiple interviewers for the same role.
Replace: [job title], [list 5 competencies].
Works in: ChatGPT, Claude, Copilot (paste into Word)

Prompt 15 — Case Study or Take-Home Brief

“Design a realistic take-home assignment for a [job title] candidate. The task should take no more than [X hours] and assess their ability to [describe core skill]. Include: a context-setting brief, clear deliverables, and the evaluation criteria we will use to score the output. Avoid tasks that require proprietary data or unreasonable time investment.”

Use this when: Adding a skills assessment stage to your interview process.
Replace: [job title], [X hours], [describe core skill].
Works in: ChatGPT, Claude

Prompt 16 — Interview Debrief Facilitation Guide

“Write a facilitation guide for a 30-minute post-interview debrief for a [job title] role. Include: an agenda, the key questions to ask each interviewer, a method to prevent groupthink (e.g., individual scores before group discussion), and a framework for reaching a hiring decision when the panel is split. Keep it practical and under one page.”

Use this when: Running a panel interview debrief with multiple hiring stakeholders.
Replace: [job title].
Works in: ChatGPT, Claude

Prompt 17 — Offer Letter Talking Points

“Write talking points for a verbal offer call for a [job title] candidate. Cover: (1) the offer summary — base salary, bonus, equity, benefits, (2) why we are excited about them specifically (leave a [CANDIDATE STRENGTH PLACEHOLDER] for me to fill in), (3) how to handle counter-offers, and (4) the timeline for accepting. Tone: enthusiastic but professional.”

Use this when: Preparing to make a verbal offer before sending the formal letter.
Replace: [job title]. Fill [CANDIDATE STRENGTH PLACEHOLDER] manually.
Works in: ChatGPT, Claude, Copilot

🚀 5. Onboarding Prompts

AI is reducing onboarding time for 56% of organisations, according to Rippling’s 2026 survey of 1,000+ HR leaders. The prompts below help you build personalized, structured onboarding communications and materials without starting from a blank page every time a new hire joins. Copilot has a particular advantage here — if you are working inside Microsoft 365, it can reference your existing onboarding documents and adapt them to each new hire’s role.

Prompt 18 — Pre-Start Welcome Email

“Draft a warm pre-start welcome email to a new hire joining as [job title] in [department] at [company name]. Their start date is [date]. Cover: (1) a genuine welcome, (2) first-day logistics — where to go, what to bring, who to ask for, (3) what their first week will look like at a high level, and (4) who to contact if they have questions before day one. Tone: warm, human, not corporate. Under 200 words.”

Use this when: Welcoming a new hire 3–5 days before their start date.
Replace: [job title], [department], [company name], [date].
Works in: ChatGPT, Claude, Copilot (in Outlook)

Prompt 19 — 30-60-90 Day Onboarding Plan

“Create a 30-60-90 day onboarding plan for a new [job title] joining [department] at [company name]. For each phase, include: learning priorities, key relationships to build, deliverables expected, and how success will be measured. Make it role-specific — not generic HR boilerplate. [Include any context about the team, current priorities, or tools they will use.]”

Use this when: Setting clear expectations for a new hire’s first quarter.
Replace: [job title], [department], [company name]. Add team context in the final bracket.
Works in: ChatGPT, Claude, Copilot

Prompt 20 — Onboarding Checklist

“Build a comprehensive onboarding checklist for a new [job title] at [company name]. Organize into three sections: Before Day 1 (IT setup, access, welcome email), Week 1 (introductions, systems training, culture orientation), and First 30 Days (role-specific learning, first deliverable, manager check-in schedule). Format as a checklist with checkboxes. Include a ‘Who is responsible’ column: HR, Manager, or IT.”

Use this when: Creating a reusable onboarding checklist for a specific role or department.
Replace: [job title], [company name].
Works in: ChatGPT, Claude, Copilot (paste into Word)

Prompt 21 — Welcome Email Sequence (Multi-Touch)

“Write a 3-part welcome email sequence for a new hire joining [department] at [company name]. Email 1 (Day 1): first-day encouragement and logistics reminder. Email 2 (End of Week 1): check in on how the first week went and prompt them to ask any questions. Email 3 (End of Month 1): acknowledge their progress, confirm next steps. Each email under 120 words. Tone: warm, genuine, not scripted.”

Use this when: Building an automated or scheduled onboarding email sequence.
Replace: [department], [company name].
Works in: ChatGPT, Claude, Copilot

Prompt 22 — Manager Preparation Guide for New Hire

“Write a one-page guide for a hiring manager preparing to onboard a new [job title]. Cover: (1) what to do before day one, (2) how to structure the first 1:1, (3) common onboarding mistakes to avoid, and (4) how to set the new hire up for a strong first 90 days. Tone: practical, peer-to-peer, not preachy. Under 400 words.”

Use this when: Equipping hiring managers with structured onboarding guidance.
Replace: [job title].
Works in: ChatGPT, Claude, Copilot

📊 6. Performance Management Prompts

Performance review writing is one of the most time-consuming and emotionally draining tasks in HR. It is also where AI adds the most immediate value — turning rough manager notes into balanced, specific, legally sound written feedback in under two minutes. The critical rule here: never paste an employee’s name, job title, salary, or performance rating into a public AI tool. Use role descriptors and anonymized notes only. For guidance on AI data handling in HR contexts, our DPIA for AI Systems guide explains exactly what triggers a mandatory data protection impact assessment.

Prompt 23 — Turn Notes into Performance Review Feedback

“Turn these manager notes into structured, balanced performance review feedback for a [job title — no name]. The feedback should: (1) open with a specific strength and supporting example, (2) identify one area for development with a concrete suggestion, and (3) close with forward-looking expectations. Keep it under 250 words. Use professional but plain language — no HR jargon. [Paste anonymized notes below.]”

Use this when: Converting rough manager notes into formal written review feedback.
Replace: [job title]. Paste anonymized notes — no employee names or identifiers.
Works in: ChatGPT, Claude (best for nuanced tone), Copilot

Prompt 24 — Difficult Feedback Script

“Script a difficult feedback conversation for a manager addressing [describe the performance issue — no names]. Use the SBI framework (Situation, Behavior, Impact). Include: (1) how to open the conversation, (2) the SBI statement, (3) how to respond if the employee gets defensive or shifts blame, and (4) how to close with a clear next step. Include phrases to use and phrases to avoid.”

Use this when: Preparing a manager for a difficult performance conversation.
Replace: [describe the performance issue]. Keep anonymized — no names.
Works in: Claude (strongest for sensitive language), ChatGPT, Copilot

Prompt 25 — Performance Improvement Plan (PIP) Draft

“Draft a Performance Improvement Plan (PIP) outline for a [job title] role. Include: (1) the performance gap being addressed, (2) specific, measurable improvement targets, (3) a timeline — typically 30, 60, or 90 days, (4) support resources the company will provide, and (5) the consequences if targets are not met. Treat this as a first draft requiring legal and HR review before use. [Describe the performance issue in general terms — no names.]”

Use this when: Drafting the framework for a formal PIP process.
Replace: [job title]. Describe the issue generally. Always have legal review the output.
Works in: Claude, ChatGPT, Copilot

Prompt 26 — 360 Feedback Survey Questions

“Write 10 questions for a 360-degree feedback survey targeting [job title or seniority level — e.g., mid-level managers]. Mix: 5 rating-scale questions (1–5) and 5 open-text questions. Cover these competency areas: [list 4–5 areas, e.g., communication, decision-making, team development, accountability]. Ensure all questions are behaviorally anchored — no vague questions like ‘Is this person a good leader?’”

Use this when: Designing a 360 feedback cycle for a specific role level.
Replace: [job title or seniority level], [list competency areas].
Works in: ChatGPT, Claude, Copilot

Prompt 27 — Goal-Setting Framework for Annual Review

“Help me write 3 SMART goals for a [job title] for the next [review period — e.g., Q4 2026 / full year 2027]. Each goal should: be specific and measurable, align with the team priority of [describe team priority], include a clear deadline, and specify how success will be measured. [Include any context about the role’s current focus or challenges.]”

Use this when: Setting structured goals at the start of a review cycle.
Replace: [job title], [review period], [describe team priority]. Add role context.
Works in: ChatGPT, Claude, Copilot

📝 7. HR Policy Drafting Prompts

Policy drafting is where Claude Sonnet 4.5 and GPT-5 have a structural advantage over Copilot — they handle long, compliance-sensitive documents with greater care and flag areas requiring legal review more reliably. Every policy prompt in this section includes a mandatory instruction to flag anything requiring legal or employment lawyer sign-off. Never publish AI-drafted policy without that review, particularly in jurisdictions with specific AI employment regulations such as the EU AI Act’s requirements for high-risk AI in HR contexts — covered in detail in our EU AI Act Explained guide.

Prompt 28 — AI Use Policy for Employees

“Draft a plain-English AI use policy for employees at [company name], a [company size and industry] company. Cover: (1) approved AI tools employees may use, (2) what data must never be entered into AI tools (client data, personal data, confidential IP), (3) how to handle AI-generated outputs — review requirements before publishing or sending, (4) prohibited uses (e.g., AI-generated performance assessments without human oversight), and (5) how violations are handled. Flag any sections requiring legal review. Under 600 words.”

Use this when: Creating or updating a company AI acceptable use policy.
Replace: [company name], [company size and industry].
Works in: Claude (recommended for policy drafts), ChatGPT, Copilot

Prompt 29 — Remote Work Policy

“Draft a plain-English remote work policy for [company name], a [company size] [industry] company. Cover: eligibility criteria, core hours expectations, equipment and expense reimbursement, data security requirements for home working, performance expectations, and the process for requesting remote work. Flag anything a lawyer should review — particularly around state-specific employment law. Under 600 words.”

Use this when: Creating or updating your remote work policy.
Replace: [company name], [company size], [industry].
Works in: Claude, ChatGPT, Copilot

Prompt 30 — Disciplinary Procedure Outline

“Outline a fair, legally defensible disciplinary procedure for [company name]. Cover: informal warnings, formal written warnings, final written warnings, suspension, and termination. Include: who must be present at each stage, the right to be accompanied, appeal rights, and documentation requirements. Flag any step requiring legal review for [US state / UK / EU jurisdiction]. Treat this as a first draft only.”

Use this when: Drafting or updating a disciplinary procedure for the first time.
Replace: [company name], [jurisdiction].
Works in: Claude (recommended), ChatGPT

Prompt 31 — Workplace Investigation Outline

“Design a workplace investigation plan for [type of allegation — e.g., harassment complaint, misconduct, data breach] at [company name]. Include: (1) initial steps within 24 hours of the report, (2) a list of people to interview and questions to ask each, (3) a document and evidence checklist, (4) a timeline for completing the investigation, and (5) a template for the investigation report. Keep names out of this prompt. Flag anything requiring employment lawyer review.”

Use this when: Planning a formal workplace investigation.
Replace: [type of allegation], [company name]. No real names in the prompt.
Works in: Claude (strongly recommended for sensitive matters), ChatGPT

Prompt 32 — Redundancy Consultation Script

“Write a script for a manager to use in an individual redundancy consultation meeting. The script should: (1) open by explaining the business reason for the redundancy clearly, (2) explain the selection criteria used, (3) confirm the employee’s right to be accompanied, (4) outline the consultation process and timeline, and (5) explain the support available — redundancy pay, notice period, outplacement. Tone: clear, calm, compassionate. Flag any jurisdiction-specific steps for [country/state] that require legal confirmation.”

Use this when: Preparing managers for individual redundancy conversations.
Replace: [country/state].
Works in: Claude, ChatGPT

Prompt 33 — Internal Announcement for Org Change

“Write an internal announcement for [company name] employees about [describe the organizational change — e.g., team restructure, new leadership, office closure]. The message should: acknowledge that change is unsettling, explain the reason for the change clearly and honestly, describe what will and will not change for employees, and tell them what happens next. Tone: transparent, calm, human. Under 300 words. Written in the voice of [CEO / HR Director / People Team].”

Use this when: Communicating a significant organizational change to your workforce.
Replace: [company name], [describe the change], [CEO / HR Director / People Team].
Works in: ChatGPT, Claude, Copilot (in Outlook or Word)

💬 8. Employee Engagement and Communication Prompts

Employee engagement is where AI has the lowest adoption in HR — only 14% of companies use AI for employee experience tasks (SHRM, 2026). That gap is an opportunity. The prompts below help HR teams design better surveys, communicate more clearly during difficult periods, and build structured feedback loops with far less effort. These prompts work across all three tools, though Copilot has an advantage if you are distributing via Microsoft Teams or Outlook.

Prompt 34 — Employee Engagement Survey Design

“Design a 10-question employee engagement survey for [company name]. Include: 6 Likert-scale questions (1–5 Strongly Disagree to Strongly Agree) covering engagement, manager effectiveness, belonging, and growth opportunity. Add 4 open-text questions that invite honest, specific feedback. Avoid leading questions. Flag any question that might generate legally sensitive responses. Include an intro statement explaining how responses will be used and kept confidential.”

Use this when: Designing or refreshing an annual or pulse engagement survey.
Replace: [company name].
Works in: ChatGPT, Claude, Copilot

Prompt 35 — Survey Results Summary for Leadership

“Summarize these employee survey results for a leadership presentation. Identify: (1) the top 3 strengths employees highlighted, (2) the top 3 concerns or frustrations, (3) any themes that appear across multiple open-text responses, and (4) 3 recommended next steps the leadership team should consider. Present as an executive summary under 300 words. [Paste anonymized, aggregated survey results below — no individual names or identifiable responses.]”

Use this when: Preparing a leadership briefing after survey results are collected.
Replace: Paste aggregated, anonymized results. Never paste individual responses with identifiers.
Works in: Claude (best for long-text analysis), ChatGPT, Copilot

Prompt 36 — All-Hands Meeting Agenda

“Build a 60-minute all-hands meeting agenda for [company name]. Topics to cover: [list 3–5 topics]. Include: a time allocation for each section, who leads each section, and one interactive element (e.g., Q&A, live poll, breakout question). Open with energy — not housekeeping. End with a clear takeaway message employees should leave with.”

Use this when: Planning a company-wide all-hands or town hall meeting.
Replace: [company name], [list topics].
Works in: ChatGPT, Claude, Copilot

Prompt 37 — Manager Communication Guide for Difficult News

“Write a communication guide for managers delivering difficult news to their team about [describe the situation — e.g., budget cuts, missed targets, team restructure]. Cover: (1) how to prepare for the conversation, (2) what to say and what not to say, (3) how to handle emotional reactions, and (4) what to communicate in the following days to maintain trust. Keep it practical — under 400 words. No names or identifying details.”

Use this when: Equipping managers to communicate bad news clearly and compassionately.
Replace: [describe the situation]. Generalize — no names.
Works in: Claude, ChatGPT, Copilot

Prompt 38 — Recognition and Reward Announcement

“Write an internal recognition announcement for [company name] celebrating [describe the achievement — e.g., a team hitting a major milestone, a project successfully launched, an employee going above and beyond]. Tone: genuine, specific, celebratory — not corporate. Under 150 words. Include a clear statement of why this achievement matters, not just that it happened.”

Use this when: Recognizing individual or team achievement in an internal communication.
Replace: [company name], [describe the achievement].
Works in: ChatGPT, Claude, Copilot (in Teams or Outlook)

📚 9. Learning, Development and Training Prompts

Learning and development ranks third in AI adoption in HR at 17% of companies (SHRM, 2026). The prompts below help HR teams design training outlines, write learning objectives, and create manager development frameworks without needing a dedicated L&D team. These are particularly useful for small HR departments managing multiple functions simultaneously — which describes the majority of HR teams, given that HR typically represents only about 2% of total headcount (AIHR, 2026).

Prompt 39 — Training Module Outline

“Create a training module outline for [topic] aimed at [target audience — e.g., new managers, all employees, sales team]. Include: learning objectives (3–5), module structure with section titles and time estimates, one activity or exercise per section, and a knowledge check at the end. Total length: [X hours]. Format as a structured outline a facilitator could follow.”

Use this when: Designing a new training module from scratch.
Replace: [topic], [target audience], [X hours].
Works in: ChatGPT, Claude, Copilot

Prompt 40 — Manager Development Programme Framework

“Design a 6-month manager development programme framework for first-time managers at [company name]. Cover: monthly themes (e.g., Month 1: Foundations of management, Month 2: Giving feedback), one learning activity per month, one peer discussion prompt per month, and a final capstone deliverable. Keep it practical — most managers will have 1–2 hours per month for structured development.”

Use this when: Building a structured development pathway for new or promoted managers.
Replace: [company name].
Works in: Claude, ChatGPT, Copilot

Prompt 41 — Learning Objectives for a Workshop

“Write 5 learning objectives for a [duration — e.g., half-day] workshop on [topic] for [target audience]. Each objective should follow the format: ‘By the end of this workshop, participants will be able to [specific, measurable action verb + outcome].’ Use Bloom’s Taxonomy action verbs. Avoid vague objectives like ‘understand’ or ‘be aware of.’”

Use this when: Drafting learning objectives for a new workshop or training event.
Replace: [duration], [topic], [target audience].
Works in: ChatGPT, Claude, Copilot

Prompt 42 — Post-Training Evaluation Survey

“Write a short post-training evaluation survey (8 questions) for a [topic] training session at [company name]. Include: 4 Likert-scale questions on content relevance, delivery quality, and likelihood to apply learning. Add 3 open-text questions on what was most useful, what could be improved, and one thing they will do differently. Add 1 Net Promoter Score question: ‘How likely are you to recommend this training to a colleague? (0–10)’”

Use this when: Evaluating the effectiveness of a training session immediately afterward.
Replace: [topic], [company name].
Works in: ChatGPT, Claude, Copilot

🔒 10. What NOT to Put in AI Prompts — HR Data Safety Rules

This section is mandatory reading before using any prompt in this library. AI and data privacy in HR is a legally sensitive area — and the most common mistake HR teams make is not poor prompting, it is pasting personal data into public AI tools. Every major AI provider’s free and standard paid tiers process your inputs for model training unless you explicitly opt out or use an enterprise API plan. That means anything you paste may be retained.

Data Safety Rule: Never paste an employee’s name, job title combined with salary, performance rating, medical or disability information, protected characteristic data, disciplinary record, or any personally identifiable information into a consumer AI tool. This applies to ChatGPT Plus, Claude Pro, and any non-Enterprise Copilot tier.

What You Must Never Paste Into a Consumer AI Tool

• Employee full names combined with any HR data
• Salary, bonus, or compensation details linked to a named person
• Performance ratings or PIP status linked to a named individual
• Medical, disability, or mental health information
• Protected characteristic data (age, race, religion, gender, pregnancy)
• Disciplinary records or termination reasons linked to named employees
• Candidate contact information (email, phone, home address)
• Social Security Numbers, National Insurance numbers, or tax identifiers
• Any data subject to GDPR, HIPAA, CCPA, or equivalent privacy law

Safe Substitution Rules

• Replace names with job titles only: “Senior Engineer” not “James Chen”
• Replace specific salary figures with ranges: “£60,000–£70,000 band” not “£64,500”
• Replace performance ratings with descriptors: “meeting expectations” not “3/5”
• Aggregate survey data before pasting — never paste individual responses
• Use enterprise tiers (ChatGPT Enterprise, Claude for Enterprise, M365 Copilot) for any workflow involving real employee data — these tiers do not use your inputs for training

Prompts 43–50 — Quick-Use Single-Line Prompts

The following eight prompts are intentionally short — designed for situations where you need a quick output without a detailed setup. Copy, replace the brackets, submit.

Prompt 43 — Reference Check Questions: “Write 8 reference check questions for a [job title] candidate. Mix: 3 questions about their work style, 3 about specific skills ([list skills]), and 2 open questions that invite the referee to volunteer any concerns.”

Prompt 44 — Exit Interview Questions: “Write 10 exit interview questions for a [job title] leaving [company name]. Make them open, non-leading, and designed to surface honest feedback about management, culture, and growth opportunities. Include one question about what would have made them stay.”

Prompt 45 — Counter-Offer Response Script: “Write talking points for a manager responding to an employee who has received an outside offer. The manager wants to retain them. Include: how to open the conversation, what questions to ask to understand the real motivation for leaving, and how to respond if the reason is salary vs. career growth vs. management issues.”

Prompt 46 — Internship Programme Outline: “Create a 12-week internship programme outline for [department] at [company name]. Include: weekly themes, 3 structured learning experiences, 2 networking activities, a midpoint check-in framework, and a final presentation or project deliverable.”

Prompt 47 — Team Meeting Agenda (Weekly): “Write a reusable weekly team meeting agenda template for a [team name or function] team of [number] people. Keep it to 45 minutes. Include: a wins round (5 mins), blockers and priorities (20 mins), one standing L&D or knowledge-share slot (10 mins), and actions summary (10 mins).”

Prompt 48 — Salary Band Research Brief: “Write a brief I can send to a compensation consultant to research salary benchmarking data for a [job title] role in [location]. Include: the data points I need (base salary percentiles, bonus structures, equity benchmarks), the sources I should request (e.g., Radford, Mercer, LinkedIn Salary), and 3 questions to ask the consultant to validate the data.”

Prompt 49 — HR Metrics Dashboard Description: “Write a plain-English description of 8 HR metrics every people team should track. For each metric, explain: what it measures, why it matters, a typical benchmark range, and one action to take if the metric is outside the healthy range. Format as a table. Metrics to cover: [list 8 — e.g., time-to-hire, offer acceptance rate, 90-day attrition, eNPS, absenteeism rate, internal mobility rate, cost-per-hire, training completion rate].”

Prompt 50 — AI Governance Briefing for HR Leaders: “Write a 1-page briefing for senior HR leaders on the governance obligations that apply when using AI tools in HR processes in 2026. Cover: EU AI Act requirements for high-risk AI in recruitment and performance management, GDPR obligations when processing candidate or employee data through AI, and 3 practical steps an HR team should take before deploying any AI hiring tool. Flag anything requiring legal review.”

🔗 Related Articles

• 📖 Claude vs ChatGPT vs Gemini: Business Comparison & Pricing (2026)
• 📖 DPIA for AI Systems: A Step-by-Step Compliance Guide (2026)
• 📖 EU AI Act Article 27 FRIA: What It Is and Who Must Comply (2026)
• 📖 AI Governance 101: Build a Policy That Actually Works
• 📖 What Is AI Safety? A Plain-English Guide to Why It Matters (2026)