How HR Teams Use AI for Recruiting and Onboarding

A single job post can bring in hundreds of applications in a few days, many of them written with AI. Recruiters are stretched, hiring managers want shortlists yesterday, and new hires expect a smooth first week. It is no surprise HR teams are turning to AI. The question is how to use it without creating a discrimination problem.

This guide covers the practical ways HR teams use AI across recruiting and onboarding, where the legal risks concentrate, and the US hiring laws employers need to watch in 2026. This is not legal advice; work with employment counsel on your specific setup.

Where AI fits in the hiring and onboarding lifecycle

The pattern is clear. AI that helps HR write, schedule and answer is low risk. AI that scores, ranks or rejects people is where legal and fairness problems arise.

Recruiting: practical uses that save time

Job descriptions. Generative AI can turn a hiring manager’s notes into a clear job post, suggest more inclusive wording and remove unnecessary requirements that shrink the applicant pool. Review every draft so the listed qualifications are genuinely job-related.

Sourcing and rediscovery. AI search inside an applicant tracking system (ATS) can surface past applicants who fit a new role, saving sourcing time and giving strong runners-up another chance.

Scheduling and candidate communication. Scheduling assistants that coordinate calendars and answer basic applicant questions are among the safest, highest-return uses. Candidates get faster responses; recruiters get hours back.

Interview support. AI can draft structured interview guides tied to the job’s competencies and summarize interview notes. Structured interviews tend to be fairer and more consistent than unstructured ones, so this is a good use. Tell candidates if you record or transcribe interviews, and follow state recording consent rules.

Screening and assessments. Resume scoring, chatbot pre-screens and skills tests can handle volume, but they carry the most risk. More on that below.

Candidates are using AI too. Our guide on how to use AI in your job search shows what the other side of the table looks like.

Onboarding: where AI quietly helps

Onboarding is a strong fit for AI because it is repetitive and information-heavy:

• Personalized onboarding plans based on role, location and team
• Policy and benefits chatbots that answer “how do I enroll in the 401(k)?” at 9 p.m.
• Paperwork reminders for I-9, tax forms and policy acknowledgments, with HR still handling verification steps
• Training paths that adjust to the new hire’s role and prior experience
• Manager nudges reminding managers to schedule check-ins at 30, 60 and 90 days

The main risks here are accuracy and privacy. A benefits chatbot that gives wrong answers can cause real problems, so ground it in your actual policy documents and route complex questions to HR. Limit what employee data the tool can see. If you are setting up your first HR processes, our guide on how to hire your first employee covers the basics.

Key takeaway: Use AI freely for writing, scheduling and answering questions. Use it cautiously, with testing and human review, whenever it scores, ranks or rejects people.

The bias risk in AI hiring tools

AI screening tools learn patterns from data. If that data reflects who was hired in the past, the tool can learn to favor similar candidates, reproducing historical imbalances by sex, race, age or disability. It can also rely on proxies: ZIP codes, graduation years, employment gaps or certain schools can stand in for protected traits.

Practical warning signs:

• The vendor cannot explain which factors drive the score
• The tool was trained mostly on another company’s data or on your own past hires without adjustment
• There is no subgroup testing or bias audit available
• Applicants with disabilities have no clear way to request an accommodation or alternative assessment

Federal anti-discrimination laws, including Title VII, the Age Discrimination in Employment Act and the Americans with Disabilities Act, apply to hiring decisions whether a person or software makes them. The EEOC removed its AI-specific technical assistance documents from its website in 2025, but that did not change the underlying laws. Private lawsuits are testing vendor liability too: in Mobley v. Workday, a federal court in California allowed age discrimination claims over AI screening to proceed as a collective action in May 2025.

Our explainer on AI bias covers how bias enters models and how to test for it, including the four-fifths rule used to spot adverse impact.

US hiring laws to watch in 2026

More states are acting, and some cities are considering their own rules. Our roundup of state AI laws tracks the latest. If you hire in Europe, AI used in recruitment is classed as high-risk under the EU AI Act.

Where to start: a 6-step plan for HR teams

1. List every AI feature in your HR stack. Many ATS, HRIS and assessment platforms have added AI scoring or matching. Know what is turned on.
2. Classify by risk. Separate writing and scheduling tools from tools that score, rank or reject candidates.
3. Question your vendors. Ask what data trained the tool, which factors drive scores, what bias testing they run, whether they provide audit results and how they support accommodations.
4. Test on your own data. Compare selection rates by sex, race and ethnicity, and age where you have the data, and repeat at least annually or after major changes.
5. Keep humans in charge. Recruiters and hiring managers should review shortlists and make final decisions, with the authority to override the tool. Document the reasons for decisions.
6. Give notice and set policy. Tell candidates and employees where AI is used, as required in NYC and Illinois and good practice everywhere, and add HR rules to your AI acceptable use policy. Add HR tools to your company’s AI governance checklist so each has an owner.

Conclusion

AI in HR is most useful for the high-volume, low-judgment work of recruiting and onboarding: writing posts, scheduling, answering questions and organizing paperwork. It becomes risky when it decides who moves forward. Employers that keep humans responsible for hiring decisions, test tools for bias, give clear notice and follow the growing set of state and city laws can save real time without trading away fairness or compliance.

Sources

• NYC Department of Consumer and Worker Protection: Automated Employment Decision Tools FAQ — Local Law 144 bias audit and notice requirements, enforced from July 5, 2023
• Illinois General Assembly: Public Act 103-0804 (HB 3773) — AI in employment amendments to the Illinois Human Rights Act, effective January 1, 2026
• California Civil Rights Council: regulations on automated-decision systems in employment — effective October 1, 2025, with 4-year record retention
• eCFR: 29 CFR Part 1607, Uniform Guidelines on Employee Selection Procedures — adverse impact and the four-fifths rule
• Colorado General Assembly: SB 26-189 — automated decision-making technology law with core duties from January 1, 2027