What should an HR team do first with Ai?

Ask your team this question. What do we do every week that takes real time and follows the same steps each time?

What is a good first job?

Good first jobs look like this. Drafting a first version of a job post. Summarizing a long policy into plain language. Turning meeting notes into a checklist. Sorting the questions that come in again and again.

Bad first jobs look like this. Screening applicants on your own. Deciding who gets an interview. Writing a final answer on pay, leave, or a termination. Those are decisions about people. A human owns them, and Ai only supports them.

Write the problem down in one sentence. If you cannot name the problem, you are not ready for the tool. That sentence is also how you will know if the tool worked. Did it save time? Did the answer get better? Did anyone have to fix it after?

What rule comes before the tool?

Before anyone uses a tool, write a short acceptable use policy. It does not need to be long. It needs to be clear.

Say what the tool may do. Say what it may never do. Say what never goes into it. Names, pay details, medical notes, and anything you would not want on a shared screen do not go into a general Ai tool.

Say who checks the output. Every Ai draft is a first draft. Someone on your team reads it, fixes it, and owns it before it goes out.

This is also where state law comes in. Some states already regulate Ai in hiring, and the rules are not the same everywhere. California, Illinois, Texas, New York City, and Colorado each took a different path. Know which tools you use, what they decide, and what records you keep. Audit work runs alongside your counsel.

How do you pick the tool and start the team?

Pick one tool the whole team can use the same way. Show them how to use it on the one problem you picked. Give them three examples of a good prompt and three examples of a bad one. Then give them one hour to try it on real work, with a person nearby to check the output.

That hour matters more than a training deck. People learn Ai by using it on work they already know, not by watching slides about it. When they see it save twenty minutes on a real task, they will use it again. When they see it get something wrong, they will learn where the line is.

Why keep a list of every tool?

Write down every Ai tool that touches HR work. The name, what it does, who uses it, and what data it sees. Update the list when a tool changes or a new one shows up.

This list is not busywork. It is the answer to the first three questions anyone will ask you. What tools do you use? What do they decide? What records do you keep?

An Ai risk and compliance audit is one way to build that list properly, with someone checking it against the rules that apply to your business.

What should you not do?

Do not let Ai write your handbook on its own. Do not paste employee information into a free tool to see what happens. Do not copy a policy from another company and change the name. Do not buy a system because a vendor promised it would fix hiring.

And do not chase every new tool. The tools change every month. The rule does not. Clear problem, plain language, human check, honest record.

What to do next

Pick the one repeat job. Write the one sentence. Then read the Why Ai page for the thinking behind the sequence, or talk to us about HR Ai upskilling, which starts with live workshops built around your team's tools, not a generic deck. FYHR Pro is free with an account, no card required, and a safe place to practice the first step before you roll it out to the team.

HR guidance, not legal advice.