3 min read

4 things you might not know about AI and recruitment 

Some surprising trends to get your teeth into.
Share on linkedin
Share on twitter
Share on facebook

What this article gets at: AI is impacting recruitment in a multitude of ways. Here are four lesser-known effects that non-experts might not know.

An interesting fact before we get started: most candidates still find their new job via word-of-mouth.


AI can be problematic, but will also widen the net

By now, most of us now know the narrative around AI and inclusive recruitment: because historical datasets of successful candidates contain societal bias — that is, they tend to be more pale, stale and male — AI is highly at risk of picking up skewed hiring criteria. 

However, AI is also better at looking beyond traditional categorisations (such as ‘School Attended’, ‘Internships Secured’) and analysing a large number of complex factors (such as ‘Overall Career Progression’). This stands to widen the net of talent and improve how candidates who have been historically overlooked are judged. Ultimately, we can expect to see far more appointments made on the basis of merit. For women and minorities, that will only be a good thing.

AI makes it easier to review old recruitment practices

There are other benefits. Evidence shows that one woman or minority person in a pool of candidates has almost no chance of making it to the offer stage in a standard, non-AI recruitment drive. But add a second one, and their chances improve dramatically. Why? The more women and minority applicants, the less organisations view them as ‘different’.

Whilst these kinds of biases are hard to undo in humans (who wants to accuse the company recruitment panel of being sexist?) amending them in a machine doesn’t create office politics — and is, therefore, much easier.

We shouldn’t train AI to ignore gender, age, disability or ethnicity 

What? Isn’t that the opposite of what you’ve been saying? Surely this will lead to the same problems?

Well, surprisingly not. That’s because if you train algorithms to ignore these factors they tend to assume the company will be best served by having many candidates of the same kind. And thus, once again (sigh), bias is created.

The answer lies in feeding AI data that emphasises the benefits of different backgrounds — train it to find people with the right basic skills, but who are varied in terms of ethnicity, class, gender and so on. After all, the real business boon of diversity is how it offers a broad church of perspectives, giving organisations a greater breadth of knowledge and understanding.

You’ll still probably need a human touch, but not always for the reasons you think

It’s a well-trodden cliché: automate recruitment processes, but ensure that human oversight is still present to keep tabs on the AI become prejudiced. However, there are lots of reasons aside from diversity for keeping humans involved in recruitment. 

Did you know that talent is four times more likely to consider a company in the future if they are offered constructive feedback? Or that the majority of appointments are still made via word-of-mouth referral? OR how gamified skills testing might alienate older applicants who are used to simply explaining their career story over a coffee? These are the kinds of human hurdles we all will need to overcome to successfully integrate AI and recruitment.

The takeaway: be cautious of AI, but remember that it will create efficiency in the long-term if properly managed.

A question to ask yourself if your business is switching to AI: How will you ensure candidates still feel valued despite automated processes? Is the AI chatbot you’ve implemented really as good as talking to a person? 

Something to read next: Juggle is crowdfunding because it wants to change the future of work. Take a look at what we are doing here.

Mya Ramanan

Get Modern Leadership delivered

Practical insights into careers and the future of work