ChatGPT Prompts for HR

Business Updated September 9, 2026 by GenPrompto
Prompt
I need to write a message to an employee about [issue, described specifically]. They're [relevant context]. Draft something [tone] that: states the specific pattern factually, acknowledges their overall standing, and opens the door for them to explain if something's going on -- end with a clear, specific next step, not a vague "let's talk." Real context: [role, specific incidents, relevant details]. Keep it under [word count].

What this does

Drafts the harder HR messages with ChatGPT -- difficult conversations, policy explanations, performance notes -- grounded in real specifics.

The best ChatGPT prompts for HR handle the parts of the job that are hardest to write cold — a difficult conversation, a policy explanation, a performance note — by giving ChatGPT the real situation instead of asking for a generic template. HR writing lives or dies on tone, and tone only works when it’s grounded in specifics.

What this produces

Drafts for the harder side of HR communication: difficult feedback conversations, policy explanations employees will actually read, and performance documentation that’s factual without being cold. Different from a recruiting or onboarding tool — this is for the ongoing people-management side of HR, after someone’s already hired.

Variations

Three ready-to-copy versions for common HR writing tasks.

Difficult Conversation Version
I need to write a message to an employee about {issue, described specifically}. They’re {relevant context, e.g. “generally a good performer”}. Draft something {tone, e.g. “firm but not harsh”} that: states the specific pattern factually, acknowledges their overall standing, and opens the door for them to explain if something’s going on — end with a clear, specific next step, not a vague “let’s talk.” Real context: {role, specific incidents, relevant details}. Keep it under {word count, e.g. “150 words”}.
Policy Explainer Version
Explain this policy change to employees in plain language: {paste the real policy or summarize it}. Structure it as: what’s changing in one sentence, why (if you can share the reason), and what employees need to actually do differently, if anything. Address the question they’ll actually be thinking — {the real concern, e.g. “does this affect my current PTO balance”} — directly in the first two sentences, don’t bury it.
Performance Documentation Version
Help me document a performance conversation I just had with an employee, for the record. What happened: {real summary}. What was discussed: {real summary}. What was agreed: {real summary}. Structure it as three labeled sections: Situation, Discussion, Agreed Next Steps (with a date). Keep it factual and neutral — no editorializing, no adjectives describing the employee’s attitude or character.

Who this is for

HR generalists, people managers without a dedicated HR team, and small-business owners handling HR themselves who need help writing the harder messages — not the easy ones. If you’re specifically writing job postings or onboarding plans, AI Prompts for HR covers that hiring-stage content, and for candidate outreach specifically, see ChatGPT Prompts for Recruiters.

Example Output

Tested on ChatGPT, August 2026. Prompt: draft a firm-but-not-harsh message to a retail shift lead about repeated lateness (4 times in 3 weeks, shifts start 9am), opening the door for them to explain.

Output:
“Hi [Name], I wanted to address your recent lateness. Your shift starts at 9:00 AM, and you’ve been late 4 times in the past 3 weeks. I appreciate your overall performance, but as shift lead, being on time is important for the team and store operations.

Please make sure you’re ready to start your shift at 9:00 AM going forward. If there’s something affecting your ability to arrive on time, please let me know so we can discuss it.”

That’s a genuinely usable first draft — factual, not accusatory, and it does open the door — but it reads a little terse for a “firm but not harsh” request. It’s a good example of a draft to warm up, not send as-is; see the tip below on tone specificity.

Tips for better results

  • If the first draft reads colder than you wanted, ask for a specific adjustment (“add one sentence acknowledging their usual reliability before the ask”) rather than just “make it warmer,” which tends to produce vague softening instead of a real fix.
  • For performance documentation, explicitly ask for neutral, factual language — without that instruction, drafts can drift toward editorializing (“employee seemed unmotivated”) that has no place in a formal record.
  • For policy explanations, name the real employee concern upfront — a policy explained in the order it was written internally rarely matches the order employees actually care about it.
  • Always review before sending — these are strong first drafts, not final messages, especially for anything that could become part of a formal record.

What didn’t work as well

Asking for “a message about lateness” with no real details produces a draft so generic it could apply to any employee at any company — no specific incident count, no specific role, nothing that signals you’re addressing this person’s actual situation rather than running a template. It reads exactly like what it is: a form letter. Naming the real role, the real pattern (4 times in 3 weeks, not just “sometimes late”), and the real tone you want is what makes the draft sound like it was actually written for this conversation.

FAQ

Should I ever send an HR message straight from ChatGPT without editing?

No — treat every output here as a strong first draft. Read it in your own voice before sending, and for anything going into a formal record or that could have legal weight, have it reviewed by whoever handles HR compliance at your company.

Can ChatGPT help me figure out if a situation needs formal disciplinary action?

It can help you think through the situation, but employment law and disciplinary processes vary significantly by location and company policy — consult your actual HR policy or legal counsel for anything with real consequences, not just a chat response.

How much personal detail about the employee should I include?

Include only what’s relevant to the situation (role, pattern, specifics) — avoid pasting in anything you wouldn’t want stored outside your HR system, since ChatGPT conversations aren’t a secure HR record-keeping tool.

What if the employee’s explanation changes what I need to write?

Go back to the same conversation and ask ChatGPT to revise based on the new context — it’s much faster than starting over, and keeps the tone consistent with your first draft.

Related