AI can take a large share of routine HR work off your team: drafting job descriptions, scheduling interviews, summarising CVs against clear criteria, answering policy questions from your handbook and guiding new joiners. It should not decide who gets hired, promoted or let go. Hiring is where AI mistakes hurt people and create legal risk, so the right model is AI that prepares and a person who decides, with written criteria, bias checks and careful handling of personal data. Below we compare traditional HR with AI used wrong and AI used right, and give you the rules to use AI in hiring safely.
How are HR teams using AI today?
Use is growing fast. SHRM's 2025 Talent Trends research found that 43% of HR professionals use AI for HR tasks, up from 26% the year before. According to SHRM's summary of its State of AI in HR 2026 report, recruiting leads HR adoption at 27%, seven of the top ten HR uses of AI are in recruiting, and fewer than half of organisations have formal AI governance policies.
In other words, AI is moving into the most sensitive part of HR faster than the rules that should govern it.
How does traditional HR compare with AI-assisted HR?
| HR task | Traditional way | AI used wrong | AI used right |
|---|---|---|---|
| Job descriptions | Copied from an old posting | Generic AI text with requirements nobody needs | AI drafts from the role's real tasks and HR removes biased or inflated wording |
| CV screening | A recruiter reads every CV | A tool ranks and rejects candidates on its own | AI summarises each CV against written criteria and a person decides |
| Interview scheduling | Endless emails and calls | Automated messages that ignore candidate replies | AI proposes slots and books them and the recruiter handles exceptions |
| Interview notes | Handwritten and scattered | Recording candidates without telling them | With consent AI summarises against the scorecard and interviewers confirm |
| Onboarding | A thick file and many repeated questions | A public chatbot answering from the internet | An assistant answers from your approved handbook and points to HR when unsure |
| Performance reviews | Written once a year from memory | AI writes the rating | AI gathers goals and feedback into a draft and the manager writes the judgement |
Why is AI in hiring risky?
Because a model learns from past data, and past hiring data often carries past bias. The best-known case is Amazon. In 2018 Reuters reported that Amazon had scrapped an experimental AI recruiting tool after finding it penalised CVs that included the word "women's", as in "women's chess club captain". The tool had learned from ten years of CVs submitted mostly by men.
The second risk is human. Saudi Arabia's data and AI authority, SDAIA, warns in its AI Bias Reference Guide about automation bias: the tendency of people to favour outputs generated by machines over human judgement. A recruiter who sees a ranked list tends to trust the order, even when it is wrong.
Regulators treat hiring AI as high risk. New York City's Local Law 144 has required, since enforcement began in July 2023, that automated employment decision tools have a bias audit within one year of use, a public summary of the results and notice to candidates. The EU AI Act classifies AI used for recruitment and selection as high-risk, and after the 2026 amendments those obligations are set to apply from December 2027.
Where can AI help HR safely?
- Writing. Job descriptions, offer letter drafts, policy summaries and internal announcements, reviewed by HR.
- Scheduling and logistics. Interview booking, reminders and onboarding checklists.
- Answering employee questions. Leave rules, benefits and procedures, answered from your approved handbook only, with a clear route to a person.
- Summarising. CVs against written criteria, interview notes against the scorecard and themes from exit interviews or engagement surveys.
- Training. Turning policies into short guides and quizzes for new joiners.
What should HR never hand over to AI?
- The decision to hire, reject, promote, discipline or dismiss anyone.
- Performance ratings and pay decisions.
- Sensitive employee data in public AI tools: salaries, medical information, disciplinary records and ID documents.
- Judging candidates on video, voice or "personality" signals that cannot be explained.
What do data protection laws in the region mean for HR AI?
Employee and candidate data is personal data. Saudi Arabia's Personal Data Protection Law came into force on 14 September 2023, and Egypt passed its Personal Data Protection Law No. 151 of 2020. In practice, that means knowing where your HR tool stores data, limiting what goes into AI to what each task needs, and never pasting candidate or employee files into public chatbots. Check the details for your situation with a legal adviser.
What rules should you follow when using AI in hiring?
- Write the selection criteria before you open any CV, and use them for every candidate.
- Let AI summarise and highlight, not rank and reject.
- Have a person review every rejection, not only the shortlist.
- Test the tool regularly: compare its output for similar profiles from different groups.
- Tell candidates when AI is used and how a person stays involved.
- Keep a log of what the tool suggested and what the recruiter decided.
- Ask vendors where data is stored, whether it trains their models and whether they have run a bias audit.
Our AI governance policy for SMBs includes a simple approval matrix you can adapt for HR.
How do you know AI is helping HR?
- Time to hire and time to schedule interviews.
- Recruiter hours per hire.
- Share of employee questions answered without a ticket, and the accuracy of those answers.
- Diversity of shortlists compared with applicant pools, checked regularly.
- New joiner time to productivity and early turnover.
This article is part of our series The AI-Ready Company. Read the opening article, Traditional vs AI: what really changes and why most companies use AI wrong, and our plan for training your team on AI.
Want an HR assistant that answers from your own policies and keeps every decision with your people? Tell us how your HR team works today. Talk to us.
