How the US combats bias in AI-assisted recruitment

AI is gaining traction in the recruitment scene, but can it overcome the challenges of inherent bias?
30 December 2019

AI is taking new steps for fairer and better recruitment. Source: Shutterstock

The recruitment scene has a few leading technologies that transformed the hiring and training process, as we know today.

From fast-food giant McDonald’s deploying voice-based technology for candidates to apply for roles in local chains, to Walmart’s use of virtual reality (VR) to access the abilities of potential candidates, recruiters have witnessed the role of artificial intelligence (AI) in leveraging the hiring and selection procedure. 

Even though technological advancement has brought new opportunities and changes in the landscape of recruitment, several experts are cautions towards the inherent bias of AI in recruitment

The concern is not without reason. To put things in context, AI is an algorithm that people train based on available data. In human resources, available data includes past job postings, applications received, and successful hires.

What happens when these data are ‘fed’ into the algorithm? It will learn and repeat the processes that result in ‘successful hires’. As a result, AI replicates biased thought processes that are, at times, conscious or subconsciously made by people. 

Even so, there is no slowing down the digital transformation taking place in human resources, especially in the US, as a study revealed 73 percent of CEOs and CHROs in the US plan to use more AI in the next three years to improve talent management.

The survey titled, Talent Intelligence And Management Report 2019-2020was conducted by Harris Interactive in collaboration with Eightfold and compiled of 1,350 CEOs and CHROs from the US, UK, Germany, and France. 

Moreover, the adoption of AI for recruitment in the US doubled from 22 percent in 2018 to 47 percent in 2019. To combat AI bias in recruitment, an effective practice, known as “candidate masking“, is gaining momentum in the US recruitment process. 

Candidate masking is the use of AI to hide a candidate’s details and a majority (79 percent) of organizations reported it had improved the recruitment process by tackling the issue of bias in AI.

Hence, organizations are presented with this scalable and effective approach to minimize the impact of bias in recruitment. 

Based on the study, HR teams in the US are taking the lead in prioritizing and improving digital skills to bolster the hiring and selection process with new technologies. Moreover, US-based CEOs and CHROs are also the most likely to recognize when they are behind the curve on technology. 

In the US, acquiring technology skills and adopting new technologies are key to address the challenges in recruitment, such as attracting candidates, retaining talent, and improving the candidate’s experience. 

In addition to that, enterprises in the US have already integrated AI in automating repetitive tasks (44 percent) and improving employee retention (42 percent). 

Recognizing the value of AI in recruitment, the US is set on a rigorous adoption of AI to address the talent crises while taming the bias beast with technological solutions. 

However, 43 percent of the US workforce still faces discrimination and more than one-third of employees highlighted the negative effects of bias. 

Presently, IBM researchers are looking into bias-detection technologies whereby AI-driven developments are trained to replicate anti-bias thought processes used by people when making decisions. 

Simultaneously, developments towards “Explainable AI,” a set of tools and framework that allows AI adopters to monitor how machine learning models make decisions. These initiatives aim to eliminate biases in the recruitment process through stricter regulations and monitoring of technologies.