Air Methods

Artificial intelligence has become difficult to avoid in conversations about the future of HR. For talent acquisition teams, the promise is especially appealing, identify qualified candidates, reduce administrative work, schedule interviews and give recruiters more time to focus on the people we want to hire.

The pressure to adopt these tools, however, can arrive before we have fully defined the problem we need them to solve. A product demonstration can show us what technology can do. To understand whether it will improve our hiring process, we need a closer look at our organization, our candidates and the work that still needs human attention.

There are real benefits. SHRM’s 2025 Talent Trends research found that 89% of HR professionals whose organizations used AI for recruiting reported time savings or increased efficiency. That is a compelling reason to explore its potential. It does not, by itself, tell us whether an individual organization will fill positions faster or make better hiring decisions. SHRM, 2025 Talent Trends

We also need to be clear about what we are evaluating. Some recruiting capabilities use AI; others, such as moving an applicant to the next step after predetermined requirements are met, may be ordinary workflow automation. Understanding that distinction helps us ask better questions about the product and the oversight it needs.

Measuring the Work AI Actually Removes

The first question I would ask is whether we are reducing the total workload or simply moving it.

AI tools are entering the hiring process from both directions. Employers are using them to review applications, while candidates can use them to prepare and submit applications. Ashby’s 2026 Recruiter Productivity report found that applications per hire remained above 300 throughout 2025 in its platform data. The company also reports hearing from recruiting teams about increases in applications submitted through automated tools. These findings illustrate the volume challenge, although they do not establish how much of the increase AI caused. Ashby, 2026 Recruiter Productivity Report.

“The value of AI should be visible in the quality of your recruiters' work and in the experience of people considering a future with your organization.”

That reflects a concern I see in recruiting, more applications do not necessarily make it easier to identify qualified candidates. When resumes begin to read like variations of the same document, someone still needs to understand the experience behind the language.

A tool may speed up the initial review while leaving recruiters with additional verification, follow-up and exceptions to resolve. Those activities should be included in the time-saved calculation.

My perspective comes from air medical organizations and the corporate roles that support them. Within that environment, different positions present different recruiting challenges. A process designed for a large applicant pool may be poorly suited to a specialized role with relatively few qualified candidates.

Designing Technology Around the Candidate

That is why you should understand the candidate population before selecting the technology.

We recently received a pitch for an AI chatbot. I could see situations in which it might be useful. But I kept returning to the candidate’s immediate goal, finding the right position and applying.

If someone already knows what they are looking for, does a chatbot help them get there? Or does a box appear in front of the page they are trying to navigate? If the candidate needs assistance, a chatbot could be valuable. If we require interaction before they can complete a straightforward task, we may be adding another obstacle.

The same consideration applies to video screening. Before making it the first step for every applicant, we should understand who will complete it, how long it will take and who may drop out.

Research gives us reason to take that concern seriously. A 2026 working paper involving more than 3,000 applicants for technology jobs found that asynchronous audio or video interviews reduced application continuation by approximately 53% relative to a process without interview screening. Live online interviews reduced continuation by approximately 20% against the same comparison. These findings come from one experimental setting, not a universal abandonment rate, but they show how an interview requirement can change who remains in the applicant pool. Avery and colleagues, A Brave New World of Hiring.

Design matters, too. A study of 27,809 applicants across 33 companies and 11 countries found that more preparation time and opportunities to record answers were associated with more favorable reactions to asynchronous video interviews. More questions were associated with less favorable reactions. Applicants over 31 reacted particularly negatively to additional questions. This supports paying attention to differences among candidates, rather than assuming an entire generation will accept or reject technology. Tilston and colleagues, Human Resource Management.

For me, the practical lesson is to investigate the people we are trying to reach. Their roles, schedules, access to technology and expectations deserve consideration alongside age or career stage. Comfort with technology does not automatically tell us whether someone wants to use a particular tool during a job search.

Candidates also value having a choice. In Greenhouse’s 2026 Candidate AI Interview Report, based on a survey of 2,950 job seekers across five countries, 46% of U.S. respondents wanted the option to request a human interview instead. That finding gives employers a concrete reason to consider an alternative path. Greenhouse, 2026 Candidate AI Interview Report.

Keeping Human Judgment in the Decision

I believe human involvement remains essential, particularly when technology influences who advances. Someone needs to understand the criteria, review the output and have the authority to question it.

That does not mean human judgment is automatically superior. The same 2026 experimental working paper found that the AI assessment tool studied predicted successive employment outcomes better than the human recruiters’ assessments. For me, that reinforces the need to evaluate both technology and our existing practices. Accountability requires more than assuming either one is always right. Avery and colleagues.

Proving Value Before Scaling the Tool

Before expanding a tool across every requisition, I would start with a defined problem and a limited pilot. Measure recruiter time, including review and rework. Track application completion, candidate withdrawals, qualified candidates reaching interviews and time to fill. Ask candidates about the experience. Review the results by job type, because an overall average can conceal a process that works well for one group and poorly for another.

Most of all, decide what recruiters should do with the time the technology saves.

That time should support stronger candidate conversations, a deeper understanding of qualifications and better partnership with hiring managers and other cross functional leaders. If we reduce the recruiter’s role to moving applicants through automated steps, we risk turning talent acquisition specialists into processors.

The value of AI should be visible in the quality of your recruiters' work and in the experience of people considering a future with your organization.