AI and Recruitment: Not as Good of a Match as You Might Think

That's not a valid work email account. Please enter your work email (e.g. you@yourcompany.com)
Please enter your work email
(e.g. you@yourcompany.com)

robot

There is an element of clickbait to articles proclaiming artificial intelligence (AI) the savior of recruitment. As someone who has been on the front lines of the recruitment technology industry for more than a decade, I’m not so convinced of AI’s ability to improve the recruitment industry as a whole.

AI will change recruiting. This much is certain. But those changes aren’t necessarily good.

There are four key areas in which AI is predicted to make the greatest impact — for better or worse — within the hiring process. Let’s take a look at what we can realistically expect:

1. Shorter Hiring Times

Let’s start on a positive note: Fully operational AI will shorten hiring times, reduce hiring costs, and improve the overall quality of candidates that land in a hiring manager’s inbox. Some highly advanced AI companies are already using clever avatars that can manage millions of applications a second, all while learning from every interaction.

The pure basis of AI is to allow machine-learning inputs to further improve and refine the overall system. If the machine learning inputs are coming from interactions between machine and job seeker, you essentially have a relentless team of people in search of jobs working to educate your AI 24 hours a day. How much more perfect could the scenario get?

2. Less Bias — or Not

The most popular improvement AI is purported to bring to the recruitment industry is the elimination of human bias. We are all aware of the discrimination that occurs within society, and the hiring process is no exception. Can machines truly remove this bias from recruiting?

The short answer: No.

My thoughts are simple: Humans are the ones programming AI, whether it be through initial creation or subsequent machine learning. Because AI is programmed and taught by humans, it has the potential to learn human bias and encode it into its own operations. Even if someone were somehow able to create the most perfect, totally unbiased machine, humans are still making the hiring decisions at the end of the day.

It’s sad to say, but even with AI in recruiting, we still live in a world rife with bias and discrimination. Ending that bias is a task far more suited for humans than machines.

3. AI Knows When You L-I-E

AI can detect when a candidate is lying purely based on language analysis. In fact, some AIs are so smart they can quiz you on a subject if they believe you may be lying.

AI’s lie detecting capabilities are particularly helpful in hiring, where studies show that as many as 85 percent of candidates lie during the recruitment process. A possibly related stat shows that 46 percent of new hires fail within their first 18 months on the job.

Do we want this sort of technology embedded throughout the hiring process? Well, it would certainly improve retention figures and help companies hire the best candidate rather than the best liar. The only issue is that the language detection technology is not always 100 percent accurate. In fact, many people have managed to convince AI they are telling the truth when they are actually lying.

That said, AI could potentially go much further than simple language analysis. Some AIs have the ethically questionable ability to use heat maps to detect blood flow on a candidates face during an interview. While not a foolproof way to detect honesty, this is another tool in the repertoire of AI to locate the (supposedly) most truthful candidates. However, I don’t think the world is ready for facial scans to be the norm during job interviews.

4. Fully Operational AI Means Fewer Recruiting Jobs

I mentioned earlier that AI could one day effectively do the job of a hiring manager. If this came to pass, we might find ourselves in a world where recruiting jobs are few and far between.

There are going to be many casualties in our industry with the growth of this technology. Make no mistake. It has always taken a human to onboard another human, but one AI application could be as productive as 20 recruiting team members. Is our industry really prepared for that?

Arran James Stewart is the co-owner of blockchain recruitment platform Job.com.

By Arran Stewart