AI, recruitment, and the value of scrutiny

We’re an experienced executive research firm. And like anyone that’s spent any amount of time in this industry, we’ve seen the term ‘AI’ used vaguely and liberally by a wide range of recruitment businesses as a sales tool.

And while the technology itself continues to evolve and improve, the robustness of these claims do not.

Businesses in every sector continue to be won over by AI; its mere mention justifying big price tags.

But are the businesses making these claims really explaining how their use of AI is unique or meaningful? And more importantly, are they explaining how their use of AI goes beyond the tools we all use, every day?

AI is far more familiar than we might thin

When certain recruiters talk about their use of AI, the intention is to give the impression of cutting-edge, high-tech processes far beyond your own capabilities – or those of whatever recruitment firm you currently use.

What they don’t want you to do is scrutinise their use of it.

Take LinkedIn – the most fundamental of recruitment tools today. LinkedIn uses AI in almost everything it does.

At a very basic level, AI systems in the context of LinkedIn or other recruitment tools are powered by machine learning. And machine learning makes predictions from huge amounts of data

Data is plugged into a system, and the system ‘learns’ from that data by identifying patterns.

Those patterns can then inform tasks and decisions of every kind.

Here’s LinkedIn’s Head of AI, Deepak Agarwal, explaining the wide-ranging results AI delivers for LinkedIn users.

“In one way or another, AI powers everything at LinkedIn. We use it in ways that our members see everyday, like giving them the right job recommendation, encouraging them to connect with someone, or providing them with helpful content in the feed.?

It is at work in products for our enterprise customers, such as helping a salesperson predict the responsiveness of their leads, serving relevant advertising to our members, or helping a recruiter find new talent pools.

And AI sits underneath various other tools we’ve all become accustomed to using within the HR and recruitment space. Take Horsefly, a talent mapping and planning tool. All of its rapid and real-time candidate recommendations are made using AI.

AI powers everything at LinkedIn, Horsefly, and dozens of other systems and applications we take for granted. So if a potential recruitment partner is talking about its use of AI, make sure you ask a simple question: how is your use of AI different to what I’m already accustomed to??

Be suspicious about AI ‘removing bias’

There’s also a slightly bogus claim regularly doing the rounds that the use of AI in a recruitment process can ‘eliminate bias’.

With DEI on everyone’s agenda, this is becoming a potent claim – but once again, it’s vague and problematic.

The problem lies in the data that fuels AI.?

AI relies on aggregating huge amounts of data, but there is no guarantee that the data placed into an AI system in the first place is unbiased.

That is actually the view of the Information Commissioner’s Office (ICO), who are currently investigating whether AI systems show bias when dealing with job applicants.

They’re now asking some very important questions about this issue. Where does this data come from? And how do we know whether that data itself is free from bias?

Evidently, the ICO hasn’t yet found the proof it needs to accept claims that AI systems can remove bias. You should ask the same questions when presented with these kinds of claims.

Demand the details

None of this is to say that AI isn’t hugely valuable within recruitment. That would be an absurd argument to make – it’s already part of everything we do.

But AI is such a vast, broad and complex concept, with an infinite number of uses and possibilities.?

Simply using the term to add weight to your value proposition is at best lazy, and at worst, cynical.?

Put simply, beware of recruiters claiming to use AI in some new and unique way to bolster their sales pitch. Without specificity, it’s meaningless.?