AI-powered recruitment: A solution to or wellspring of workplace discrimination?

As uptake grows and results are published, the business case for AI-powered recruitment is only gathering strength. Understanding its shortcomings, however, is a crucial first step in protecting against discriminatory hiring practices.

Though AI-powered recruitment is still relatively new, its growth – both in terms of uptake and diversity of functions – has been astounding. From résumé screening to conducting preliminary interviews, hiring functions are increasingly being made cheaper and quicker with the adoption of artificial intelligence (AI) technologies. Let’s begin with the stats.

In terms of growth:

• The use of AI in recruitment has doubled in the past year alone.
• There are now more than 250 commercial AI recruitment tools being used by Australian businesses.
• As of last year, one in three Australian businesses reported having used AI in recruitment.
• The AI recruitment market has grown by over 64 per cent since 2020 to US$590 million – and is set to reach US$942.3 million by 2030.

Now, let’s consider the drivers behind the rapid adoption:

• It’s faster: AI can speed up hiring by up to 90 per cent.
• It’s cheaper: Early adopters have found their candidate screening costs cut by 75 per cent.
• It’s more reliable: Early adopters have seen their staff turnover decrease by up to 35 per cent, and AI-selected candidates are more likely to pass interviews.

It’s unsurprising, then, that interim chief people officer at Salesforce Nathalie Scardino said, in reference to AI recruiting: “This is not the next tech bubble – it’s already having real business implications.”

What she added, though, was that it is up to employers to ensure that AI recruiting technologies are deployed safely and sustainably. The Diversity Council of Australia (DCA), in its third instalment of the Inclusive AI at Work in Recruitment series, also expressed concerns about the ability of AI-powered recruiting to perpetuate bias and discrimination.

“We know that unless AI is deployed with a focus on diversity and inclusion, it has the potential to mirror society’s inequalities and bake in systemic biases. Conversely, if it’s used with [diversity, equity, and inclusion] front of mind, the benefits can be astounding,” explained DCA chief executive Lisa Annese.

The potential for discrimination

According to DCA, inclusive AI recruitment occurs when “AI recruitment tools automate, communicate, assess, and predict in ways that value diversity, minimise bias, and enable a diversity of talent to be hired”.

At first glance, non-human hiring might seem like a solution to discrimination and biases in recruitment rather than a contributor. According to DCA, however, this is not necessarily the case – a concerning idea when considering that 74 per cent of organisations have not taken any steps to reduce unintended bias in AI. Consider the following ways in which AI-powered recruiting can perpetuate discrimination:

• Video interview assessment tools might misinterpret or misunderstand the tone or accents of non-native speakers and/or jobseekers with unique speech patterns or who are visibly anxious.
• Job application platforms might discriminate against applicants with disabilities due to overly complex navigation, timeout restrictions, lack of video captioning and so on.
• AI-powered sourcing may prioritise efficiency and cutting costs at the expense of candidates who are harder to find. For example, women tend to be more expensive to advertise to due to a higher click-to-profit ratio.

Further, the problem is proving difficult to solve. Not only can fairness and bias be difficult to measure and, therefore, solve for, but AI governance is still in its infancy.

“The rapid advancements in AI technology have outpaced the development of comprehensive governance frameworks. However, it is important to remember that existing anti-discrimination legislation covers recruitment, and through that also AI recruitment,” said DCA.

How employers can help

According to DCA, there are five steps that employers ought to take to reduce discriminatory practices when making use of AI-powered recruitment:

1. Team up to assess diversity and inclusion impact.
2. Reflect on your readiness for inclusive AI recruitment.
3. Educate your team about bias in recruitment.
4. Acquire expertise on how bias plays out in AI recruitment.
5. Decide how to proceed inclusively with AI recruitment.