Why isn’t AI in talent acquisition improving hiring outcomes?

AI is already embedded in recruitment. It writes job descriptions, searches for candidates, screens applications, schedules interviews and supports assessments.

So where is the corresponding leap in hiring performance?

Research commissioned by ManpowerGroup Talent Solutions and conducted by Everest Group found that more than 90% of organizations have deployed AI in talent acquisition, yet fewer than 5% report transformational outcomes.

That leaves a sizeable gap between adopting AI in talent acquisition and getting better results from it.

Why isn’t AI improving recruitment outcomes?

One reason is that organizations have often applied AI to individual recruitment tasks rather than reconsidering the process around them.

Automating CV screening can process applications faster. It does not necessarily improve the criteria being used to screen them. A sourcing tool can find more candidates, but that matters less if recruiters are already dealing with too much poorly matched volume.

This distinction is becoming more important as candidates adopt the same technology. SHRM found that 85% of recruiting executives expect candidates’ use of AI to apply for jobs to increase, while 74% expect greater use during interviews.

Recruitment now has automation on both sides of the process. More activity does not automatically produce better hiring.

Is recruitment automation solving the right problem?

Much of the early case for recruitment automation was built around efficiency: fewer manual tasks, faster screening and greater recruiter capacity.

Those are useful gains. But time saved is not the same as a better talent decision.

If an organization has unclear job requirements, inconsistent assessments or disconnected recruitment technology, adding AI can simply move those weaknesses through the process faster.

This may explain why AI adoption and AI impact are moving at different speeds. The technology can automate a step without fixing what happens before or after it.

Before introducing another AI tool, talent acquisition leaders need to know what they are trying to improve. Is it time to hire? Quality? Candidate conversion? Recruiter capacity? Skills identification? Hiring manager experience?

Without that definition, measuring the value of AI recruitment becomes difficult.

What should recruiters still own?

As more administrative work is automated, the recruiter’s role does not disappear. The work that remains becomes more consequential.

Recruiters still need to understand whether a hiring requirement makes sense, challenge unrealistic expectations, interpret candidate evidence and advise hiring managers when the market does not match the brief.

SHRM’s recent analysis of the future TA function points in the same direction: recruiters are moving closer to workforce planning, internal mobility and talent strategy as AI absorbs more transactional work.

That changes what good recruitment looks like.

A recruiter who spends less time scheduling interviews but no more time advising the business has gained efficiency. A recruiter who uses that capacity to improve the hiring decision has created value.

How should enterprises approach AI in talent acquisition?

Start with the hiring problem rather than the technology.

Map where decisions are slow, where candidate quality drops, where recruiters spend unnecessary time and where hiring managers lack useful information. Then determine whether AI, process redesign, better data or a combination of the three addresses the cause.

Governance matters too. Organizations need to know where AI is making or influencing decisions, what data it relies on and where human review is required.

The aim should not be to automate as much recruitment as possible.

It should be to decide which parts of recruitment benefit from automation and where human judgment is worth protecting.

What does successful AI recruitment look like?

The useful measure of AI adoption is not the number of tools deployed or tasks automated.

It is whether hiring decisions improve.

That could mean recruiters identifying stronger candidates sooner, hiring managers receiving better market advice, candidates moving through a clearer process or TA teams spotting workforce needs before they become urgent requisitions.

AI can contribute to all of those outcomes.

But technology does not turn a fragmented hiring process into a good one simply by making it faster. The organizations that get more from AI will be the ones willing to examine the recruitment decisions underneath the automation first.