Diversity and ‘AI for HR’ – a blessing or a curse?

In this blog, we look at AI in the world of HR, focusing on one of its prime focus areas and challenges: Diversity, where the potential and the perils of AI are in hot debate.

In 2023, businesses are grappling with AI across departments and disciplines. They know it’s going to help them get ahead, and they need to adopt it now…

But exactly where and how AI can best support your business, as well as the surest ways to avoid AI pitfalls, are more obscure.

AI can be used in a variety of key ways within HR, with the potential to support diversity in your organisation, and bring the many benefits it offers.

Here are our top three (they might surprise you):


Generative AI tools like ChatGPT which specialise in crafting text can help kick off your recruiting process by creating job descriptions and ads. They can even suggest useful and relevant interview questions, or help boost your employer brand so you’re more likely to attract great candidates in the first place.

So, should you let AI hire your talent for you?

The answer is a yes / no split. You first need to ensure that your AI tools are reliably trained.

For instance, AI candidate screening may quickly identify potential, while removing the inherent bias that is often present when humans screen CVs.

If trained on the wrong data though, AI can reinforce issues of underrepresentation. One example is a quickly scrapped screening tool trained on male dominated CVs in tech. It learned, troublingly, that male candidates were simply better. As a result, women’s CV’s (identified via details like ‘women’s chess club captain’) were ranked lower, and those featuring verbs more often used by men (apparently these include ‘execute’ and ‘capture’) were prioritised.

That’s right, the AI investigated whether a cv belonged to a male or female candidate, and then didn’t want to give the woman the job…

Suffice to say, fair and equitable data sets are crucial in building non-biased, responsible AI tools. Once you’ve done that, you’ll have a truly objective screening system.

Performance reviews

Two key factors are at play here. 1: AI is completely objective, unlike humans (whether we know it or not) and 2: it sees everything.

When trying to analyse and effectively summarise an employee’s performance over 6 months or a year, there is too much potential to miss or misinterpret a lot of detail.

Many businesses may make judgments based on what stands out to the manager, rather than on comprehensive data. This runs the risk of things like proximity bias, unconscious gender or racial bias, and even language barriers becoming an unfair obstacle to progression.

AI, rather, can digest information from various sources, like project management tools, calendars and more to give a comprehensive overview of an employee’s productivity and trends in their output, creating an unbiased, meritocratic basis for review.

Learning & development

Once again, it’s about handling huge amounts of data – too large for your teams to efficiently process without the use of AI. It can identify knowledge gaps across your business and set you up to deliver training in a targeted way.

How, you ask? Here’s a great example reported by Forbes, from Miguel Caraballo, director of learning and development at Hodges-Mace:

"Imagine this, your rookie sales person is about to visit a client. When she pulls into the parking lot, the company learning-bot pings her phone offering to show a micro-learning class about positive first impressions along with a link to the prospect’s ‘About Us’ page and her manager’s last coaching notes."

In this way, AI powered tools can harness the power of delivering training that’s not only personalised, but timely. This means staying ‘in sync with what motivates or drives’ your employees. Arabelle Fedora, VP and managing principal consultant of talent management at Right Management told HR drive that this is essential to ensure lasting benefit.

Providing training in an automated way provides equal access to learning opportunities, unhindered by how you work, where you work, or how much facetime you have with your superiors. AI enhanced training adapts to employees’ requirements, skill levels and more.

This allows employees from all backgrounds with different experience levels to tap into the learning they need when they need it, supporting progression indiscriminately across your business.

What’s the takeaway?

Whichever of these areas you focus on when integrating AI to boost diversity, the difference between making it ‘a blessing and a curse’ all comes down to data!

Put simply: Feed your AI good data, you’ll get great results. Feed it biased, incomplete or unclean data, you could get a disaster. It’s the old ‘rubbish in, rubbish out’ refrain in action.

Add this to your ever-growing list of reasons that Data talent is essential for the future of your business. Act now by checking out our portfolio of data courses and apprenticeships, to equip your organisation with data expertise, AI capability, and innovative diverse talent.

Related Articles