AI for HR means using AI tools to do parts of HR work faster: drafting job descriptions and policies, scheduling and screening in hiring, building onboarding plans, summarising feedback and reviews, answering routine employee questions and analysing people data. It is strong at writing, summarising and spotting patterns. It is weak at judgements about individual people, and those should stay with a person.
Adoption is still early. In SHRM's 2026 research with more than 1,900 HR professionals, 39% of organizations had implemented AI in their HR function, and recruiting was the most common use.
This guide covers how HR teams use AI today, function by function, what to automate and what to keep human, the rules that apply, the skills an HR team needs, and a simple way to start. For ready-to-use prompts, see the companion list of ChatGPT prompts for HR.
How HR teams use AI: use cases by function
Recruiting
- Job descriptions: a first draft from the hiring manager's notes, then a check for exclusionary wording.
- Interview kits: structured questions and scorecards tied to the competencies the role needs.
- Scheduling and candidate messages: the admin that slows hiring down.
- Screening: sorting applications against stated requirements. This is the use case most covered by law, so read the rules section below before you switch it on.
Onboarding
- Onboarding plans: a task list with owners and dates for each new hire, built from your documented process.
- Welcome emails, first-week schedules and 30/60/90-day plans, drafted for the manager to edit.
- New-hire questions: answers drawn from your handbook instead of a message to HR.
Performance management
- Review forms generated for a role or a review type.
- Written summaries of self, manager and peer feedback.
- Feedback quality checks that flag vague or biased wording before a review is shared.
- Calibration prep: ratings and comments across teams in one view.
This is the area Effy AI works in. The tools are compared in performance review software.
Learning and career development
- Career frameworks: levels and competencies drafted from your job descriptions, then reviewed by managers. Platforms for running them are compared in career pathing software.
- Skills gap analysis and individual development plans; see employee development software.
Employee experience and surveys
- Survey design and, more usefully, grouping hundreds of open comments into themes in minutes.
- Action plans for managers based on their team's results.
HR operations and policy
- Policies and the handbook: first drafts and plain-language rewrites, with legal review before anything is published.
- HR helpdesk: a chatbot that answers routine questions about leave, benefits and policies from approved documents.
People analytics
- Attrition and engagement patterns, pay comparisons and headcount reporting, asked in plain language instead of built in spreadsheets.
What to automate with AI and what to keep human
A useful line: let AI prepare, let people decide.
- Automate: first drafts, summaries, scheduling, reminders, formatting, translating, answering routine policy questions, grouping survey comments.
- Assist, with a person checking: shortlists, review drafts, development plans, policy changes, analysis of pay or attrition.
- Keep human: hiring and firing decisions, ratings, promotions, pay decisions, disciplinary steps, and any conversation about someone's health, family or performance problems.
Risks of using AI in HR, and the rules that apply
Four risks come up in almost every HR use of AI:
- Bias: a model trained on past decisions can repeat them. Screening and rating tools need checking for different outcomes across groups, and AI-written reviews need a bias check like any other.
- Privacy: employee data is sensitive. Use approved tools, prefer business plans that don't train on your data by default, and remove names and identifiers when you can.
- Accuracy: AI states wrong things confidently, including about employment law and about the people in a document.
- Transparency: candidates and employees increasingly have a right to know when AI is used on them.
Law is catching up, mostly around hiring and decisions about employees:
- European Union: the EU AI Act classes AI used for recruitment and selection, and for decisions on promotion, termination, task allocation and monitoring or evaluating performance, as high-risk. After the AI Omnibus amendment, which entered into force on 27 July 2026, those obligations apply from 2 December 2027. Separately, using AI to infer employees' emotions at work has been banned since 2 February 2025, except for medical or safety reasons.
- New York City: under Local Law 144, enforced since 5 July 2023, employers using automated tools for hiring or promotion decisions need a bias audit within one year before use, must publish a summary of the results, and must notify candidates and employees 10 business days in advance.
- Illinois: since 1 January 2026, it is a civil rights violation to use AI that has the effect of discriminating in recruitment, hiring, promotion, training, discipline or discharge, and employers must notify employees when they use AI for those decisions.
- Colorado: the state replaced its 2024 AI Act with SB 26-189, scheduled to take effect on 1 January 2027, with notice and human review requirements for consequential decisions such as hiring.
This is a summary, not legal advice. Check with employment counsel before you use AI in hiring, promotion or performance decisions, especially if you hire in more than one country or state.
AI skills HR teams need
You don't need to code. The skills that make the difference are practical:
- Writing clear instructions: context, task, inputs, constraints, format.
- Organising context: keeping your policies, process documents and style in a form the AI can read, so you aren't re-explaining the company in every prompt.
- Judging output: spotting confident errors, missing evidence and biased wording.
- Data hygiene: knowing which employee data can go into which tool.
- Process thinking: writing down how a task is done today, so it can be handed to AI and checked.
How to start with AI in HR
- Write an AI use policy first: approved tools, data that never goes in, when output needs review.
- Pick one repeated task that eats time, such as onboarding plans or review summaries.
- Document how you do it today, in a page or two.
- Run it with AI for a month and compare time and quality with the old way.
- Save what works as reusable context and instructions, then pick the next task.
Effy AI's free AI for HR course follows this path with Claude, and its free skills cover tasks such as documenting your onboarding process and building a career framework.
AI tools for HR: where to look
General assistants such as ChatGPT, Claude and Copilot cover most writing and summarising work. HR software with AI built in covers the rest: recruiting platforms, HRIS, performance and engagement tools. The vendors and what their AI features actually do are compared in AI HR software.
Effy AI, our own product, uses AI to generate review forms and written summaries, with bias checks and quality nudges on written feedback. It is free for up to 5 users and starts at $3 per seat a month.
Will AI replace HR jobs?
AI is more likely to replace parts of HR work than HR roles. The most exposed work is administrative: scheduling, first drafts, routine questions, reformatting data. The least exposed is judgement and trust: coaching managers, handling conflict, deciding pay and promotions, and being accountable for decisions that the law increasingly says a person must own. HR teams that hand the first kind of work to AI get more time for the second. The wider picture beyond HR is covered in the pros and cons of AI in the workplace.
The short version
AI for HR works best on drafting, summarising and routine questions, and worst on decisions about people. Start with a policy and one repeated task, keep a person on every decision, and check the rules before you use AI in hiring.
