Drafts a one-page policy for AI in your hiring: allowed uses, uses that need a separate review (anything that screens, ranks or scores), where a person must decide, the candidate notice, data handling, and the questions to put to vendors and your legal adviser for each location you hire in.
HR operationscategory
No scriptsinstructions only, sends nothing
Any AI appupload, or copy and paste
Free, no signup
Upload the file to Claude, ChatGPT Business or Gemini Enterprise. On ChatGPT Plus, the free Gemini app or Grok, copy and paste it instead.
---
name: ai-hiring-policy
description: Drafts a one-page AI in hiring policy for an in-house team: allowed and banned uses, where a person must review, the notice candidates get, how candidate data is handled, and the questions to put to vendors and a legal adviser. Use when a recruiter or HR lead asks for an "AI hiring policy", "AI recruiting policy", "can we use AI to screen", "AI notice for candidates", "AI vendor questions", or a vendor is promising AI screening.
license: MIT
metadata:
author: Work Insiders
version: "1.0"
pack: in-house-recruiting-pack
---
# AI hiring policy
You help an in-house team write down how it uses AI in hiring, before a vendor, a candidate or a regulator asks.
If `references/company-profile.md` exists, read it for the AI tools already in use, locations you hire in, and who reviews legal questions.
## Inputs
- Every AI tool or feature touching hiring today: general assistants, ATS features, note takers, sourcing tools, assessments. Ask for this list; do not assume.
- Locations the company hires in.
## The policy
1. **Purpose.** One paragraph: AI helps us write, organise and prepare; people make hiring decisions.
2. **Allowed uses.** Examples: drafting job ads and messages, building interview kits, summarising a recruiter's own notes, scheduling, writing search strings.
3. **Not allowed without a separate review.** Anything that screens out, ranks, scores or rates candidates; facial, voice or emotion analysis; inferring traits from social media; using candidate data to train outside models.
4. **Human review.** Every decision to reject, advance or hire is made by a named person who has read the application or evidence themselves.
5. **Candidate notice.** Draft the plain-language notice: which tools, for what, that a person decides, and how to ask for an alternative or an adjustment.
6. **Data.** What candidate data may go into which tools, how long it is kept, and who can see it. Mark anything the recruiter is unsure about.
7. **Vendors.** Questions for every AI vendor: what the feature does to candidates, whether it scores or ranks, what bias testing exists and when it was last run, where data goes and whether it trains their models, and how a person can override the output.
8. **Ownership.** Who owns the policy, how often it is reviewed, and who to ask.
## Laws
Do not state what the law requires. List each hiring location and mark it `check AI hiring rules with your adviser` (some places require audits, notices or consent). Draft the questions for the person named for legal questions.
## Output
A one-page policy in plain language, the candidate notice, the vendor question list, and the adviser question list, each as its own block.
## Rules
- Every tool named comes from the recruiter's list. Do not add tools.
- Mark anything unconfirmed as `[to confirm]`.
- This is a draft for people to approve, not legal advice.
The company profile it reads (fill this in so it uses your real details)
# Company profile
Fill this in once. Every skill in the In-House Recruiting Pack reads it, so scorecards, messages, offers and policies use your company's real details. Delete any line you do not want the AI to use. Leave a line blank rather than guessing.
## Who we are
- Company name:
- Recruiter or TA lead name and title (for sign offs):
- Careers page or apply link:
- Locations and work setup (on site, hybrid days, remote):
- Headcount (roughly):
- What we make or do, in one line:
## How we hire
- ATS:
- Our standard interview stages:
- Who makes the final hiring decision:
- How fast we promise to reply to candidates at each stage:
- Do we pay for take-home work samples? (yes/no):
- Accommodation contact (name or email):
- Background or reference check policy:
## Pay and offers
- Do we post pay ranges? (always / where the law requires / never):
- Where our pay ranges come from (survey, data tool, internal bands):
- Benefits we can mention:
- Who approves an offer:
## Policies
- EEO statement:
- AI tools we use in hiring (if any):
- Who reviews legal questions (in-house counsel, outside adviser, HR lead):
## Words we never use
-
Yes. It is part of the free Work Insiders In-House Recruiting Pack, MIT licensed, so you can edit and share it.
Does AI Hiring Policy work in Gemini?
Yes. Upload it in Gemini Enterprise or drop the folder into Gemini CLI's skills folder. In the regular Gemini app, paste its instructions into a Gem.
Can I trust it with my data?
It has no scripts and sends nothing anywhere. Whatever you paste goes only to your AI assistant, under that assistant's own data terms, so remove candidate details you do not need first. It can still steer how the assistant answers, so read it before you use it.
Written by Work Insiders for in-house recruiting teams. Fill in the company profile before you use it so it never invents pay, stages or policies. People make every hiring decision. How we make money.
