The False Promises Of AI Recruitment

Can AI measure culture fit?

Artificial Intelligence-facilitated or AI-facilitated recruitment has been a buzzy concept for years now, and it was inevitable that the public launch of Chat GPT would kick that conversation into hyperdrive. Already, outlets like CNBC are touting the potential of AI to "reinvigorate the entire recruitment and retention experience," with countless companies signing up with AI recruitment companies in the hopes of streamlining and optimizing the hiring process.

It is definitely a nice idea. The hiring process, after all, is hard: filling just a single substantial role can take weeks if not months of concerted effort. For overstretched recruiters, passing the heavy lifting onto AI carries an understandable appeal. But the fact is that for competitive industries, AI does not bring much to the table—and might actually hinder recruitment efforts.

No matter how advanced the technology gets in the coming years—because when it comes to recruitment in competitive industries, the human touch is the entire game.

AI Cannot Help with Active Recruitment

Up-top, it is worth distinguishing between recruitment in high-status fields like engineering and less competitive industries. In less competitive industries, the pool of qualified applicants is much larger, and employers are routinely flooded with applications. Here, AI might make some difference, at least around the margins—sorting the wheat from the chaff and reducing the signal-to-noise ratio.

But engineers are a different story. For one thing, it is the rare engineer who actually applies to an open job listing—almost as a rule, engineers are courted. It is an active process, requiring would-be hirers to search places like LinkedIn, GitHub and AngelList and to actively identify compelling talent. 

And actually soliciting that talent—getting past the defenses of people who often hear from dozens of recruiters per week—requires a skillset that no AI can possess: a combination of old-fashioned people skills and deep knowledge of industry trends. In other words: you need to know what makes the category of professionals you are targeting tick. 

Ultimately, an AI can only really be useful when it comes to people who are actively looking for work—with the effect that relying on AI for recruitment purposes means losing 75% or more of qualified candidates. And that is not to mention the candidates you might miss through AI's well-established cultural biases. If we are going to continue to center on improving diversity, equity, and inclusion (DEI) in competitive workplaces—as we certainly should—we need to reckon with those biases before springing for AI recruitment technology.

AI and "Culture Fit"

Here is the thing, though: even if you could magically get engineers to apply to work for you without actively soliciting them, AI still would not move the needle.

Indeed, AI recruitment technology can quickly filter out candidates, who lack the relevant experience, and provide you with a stack of people who (in theory) fit the bill. This is useful — but it gets you about 10% of the way there. Because as anyone who works in recruiting knows, the crux of the hiring process is human judgment.

An AI cannot measure candidate enthusiasm. An AI cannot measure a candidate's interpersonal skills. And crucially, an AI cannot determine culture fit—which is essentially what the hiring process is all about.

"Culture fit," as a concept, is often used as a synonym for "personality"—i.e., whether a given candidate would get along with the rest of the team. But this is just one part of what culture fit means. A more holistic approach to "culture fit" would incorporate the candidate's story—the series of decisions that brought them to your company in the first place. The in-person job interview can be a charged, generative zone in which interviewers learn the kinds of things that conventional resumes cannot account for. These uncovered personal stories can be extremely useful when trying to make a hire.

Fundamentally, the problem of AI recruiting in competitive fields goes both ways. On the hiring end, recruiting personnel ends up wasting much more time on unqualified candidates, further slowing down the candidate hunt. Meanwhile, an AI recruiter deprives the candidate of a firsthand experience of what working at the company in question is actually like—because, unlike a conventional human interviewer, an AI has no personal experience of working at a given company.

Again: it is not inconceivable that an AI might help with some of the busywork of the human resources (HR) process, particularly in less competitive fields. But the substance of the hiring process—which goes far beyond filtering applications and asking simple questions—will remain the province of humans for the indefinite future.

Nancy Drees is the Founder & CEO of Vacaré Group.