Chief human resource officers (CHROs) are likely familiar with the use of generative AI (GenAI) in HR to draft job descriptions and summarize meeting transcripts. The advances offered by agentic AI to automate more HR processes have the potential to take the function to a new level—and that has ramifications for how CHROs lead their teams and manage their organizations.

GenAI performs tasks such as writing first drafts of job descriptions. Agentic AI handles not just tasks but entire workflows. It can even make decisions and act on them.

In recruitment, it could screen resumes, compile a list of top candidates, and e-mail those candidates to schedule their interviews, as The HR Observer points out. AI agents are even starting to conduct online interviews, to the chagrin of some job applicants, The New York Times reported. If interviews are conducted online (by an HR manager, or even by the AI), it can then analyze the interviews, reassess candidates, and produce a list of finalists for a recruiter or hiring manager. These are among the early implementations of HR agents. Gartner predicts that by 2028, 30% of recruitment teams will use AI agents to handle some recruitment activities.

To prepare for agentic AI in HR, experts offer CHROs five recommendations.

Personally investigate both GenAI and agentic AI

Investigating agentic and GenAI you will understand how each works and what each can (and cannot) do.

HR is usually not as tech savvy as other parts of a business, but that must change, says Shrikant Pattathil, president and CTO, Harbinger Group, a global technology company. “CHROs need to gain experience, knowledge, and confidence in how to use agentic AI. They need to understand more of the mechanics before they will truly be able to exploit the benefits,” Pattathil says.

Test out an AI chatbot to craft a memo or summarize research you have gathered (and check the results before using them). Engage an AI agent in a “conversation” to experience how a job applicant might feel.

Prepare HR employees to work with agentic AI

Part of that is education, so they understand the technology. But perhaps more important is clear, honest communication about where and how the technology could change jobs. In some cases, it could eliminate jobs, but in many cases, it could enhance work and make jobs more satisfying.

For example, thoughtful application of this technology could help ensure that the right employees are in the jobs best suited to their skills, leading to more fulfilling work and higher job satisfaction. This is a ripe area for exploration. Only 40% of the employees surveyed for a Mercer global talent trends report said that their current roles aligned with their motivations and made the best use of their skills.

Agentic AI also promises to handle more of the administrative tasks associated with work, leaving more time for strategic and creative thinking. For instance, curriculum designers spend a lot of time on things other than designing curricula, such as scheduling classes and e-mailing invitations to participants. “They are spending 90% of their time on that administrative stuff,” says Pattathil. “So, they don’t have much time to actually design the curriculum.” If AI takes care of those tasks, “they can think about how to update the curriculum, or how to personalize it for different audiences,” says Pattathil.

So far, companies don’t seem to be getting such positive messages across to their workers. In Mercer’s Workforce 2.0 Global Talent Trends 2024–2025 survey, 60% of employees said their organization wasn’t good at communicating how AI or automation will improve the way they work. Perhaps as a consequence, only 27% of employees said they expect AI or automation to improve how their job is done over the next three years.

Pick an initial HR use case for agentic AI and put it under a microscope

The AI-led interviews are one early example. CHROs are sure to have ideas here, what HR expert Josh Bersin, founder and CEO of The Josh Bersin Company, calls “a target-rich opportunity.”

“HR departments have a lot of frustrating operational areas,” Bersin says. He recommends focusing on one of your most urgent needs. Pick apart the processes and workflows in that use case, evaluating where agentic AI could be applied.

For example, in recruiting, AI could write the job description, source the candidates, and give you a shortlist of candidates it has pre-qualified, says Bersin. If you apply AI only to those tasks, “you just saved a whole bunch of time and a whole bunch of staff energy and expense of going through that whole process to get to these final candidates,” he says.

Such close examinations create opportunities to reimagine and improve the process even before applying new technology, Dan Beck, president and chief product officer, SAP SuccessFactors, points out.

“You might be shining the light of day on a process that hasn't had much attention,” Beck says. “There will be increased pressure on evaluating the existing process, asking whether it is the right process.” Also consider the specific ways that agentic AI will change specific employees’ jobs, so you can prepare staffers and train them in whatever new skills they may need.

Focus on data governance and security

Agentic AI will likely not only have access to valuable and potentially sensitive data but will also be able to decide how to use that data, which presents new security and privacy risks. CHROs need to know not only what laws currently apply but also what new regulations are developing worldwide.

For example, the European Union has passed the AI Act, which classifies the use of AI in employment-related decisions, including candidate screening and employee evaluation, as high risk. Companies that deploy high-risk AI systems in their HR activities must comply with certain provisions of the Act concerning transparency, data management, data protection, and human oversight, according to law firm Hunton Andrews Kurth LLP. This includes informing employees when AI is used and giving them a choice over whether AI systems can access their data.

Transparency and risk management extend beyond regulatory compliance. Consider the importance of protecting the company against bias claims. Why was one job candidate chosen over another, and what role did an algorithm play? How does the company ensure its processes reduce bias, if not eliminate it? As this article about ending AI bias points out, a number of organizations, from nonprofits to startups to tech giants, offer AI audit tools that can test and refine models before deployment and check for bias during their use.

Forge and strengthen relationships with the IT department

Working closely with IT not only ensures the best governance and data security but also lays a pathway to managing a joint workforce in which employees and AI agents work in tandem. AI brings HR and IT functions closer than ever, so cooperation is vital. Good communication and working relationships with IT will be key to effective, efficient, and secure deployments in HR initially, and ultimately, company-wide.

In fact, vaccine maker Moderna recently merged its HR and IT departments under its CHRO, Tracey Franklin. Franklin told The Wall Street Journal that the move was a natural evolution because she had worked with the company’s former CIO to redesign work related to the right combination of employees and technology systems.

Some experts have started describing IT as the manager of digital employees, aka AI agents. At Mobil World Congress in early 2025, a panel discussed how IT could become “virtual human resources departments,” ZDNet reported. “Instead of an HR for humans, now you have an IT department that's acting as HR for all these agents,” says panelist Fred Devoir, global head of solution architecture for telecommunications at chip maker Nvidia.

With the right preparation, CHROs can be partners with IT in applying agentic AI and can make sure employees get what they need in the process. “HR leaders have the opportunity to become ‘stewards of humanity,” write authors of the Mercer survey, “ensuring a seamless integration between AI and human capabilities,” that leaves employees better off.

Read more