Are you confident that you can evaluate a candidate's CV fairly and accurately? Using AI can eliminate human bias and make more accurate decisions about a candidate's future performance.
Accessibility Links

Can AI democratize your hiring process?

Date: 17 June 2019

View of the back of a head of man raising his hand. He is wearing a white shirt. n at a conferenceHow can HR professionals predict the future performance of a candidate? The answer, as always, is to look at past behaviour. The way we continue to define and measure success, however, continues to rely on antiquated indicators, such as an university degree or work experience, presented in an increasingly outdated format: the resume.

Research by CareerBuilder shows that most employers spend less than two minutes reading a CV, and even this may be a generous estimate. When a hiring manager has so little time to evaluate whether someone is fit for the role, this leaves little room for a candidate who doesn’t tick prescriptive boxes, or who may not have had a conventional education or career journey.

This is an opportunity cost that neither candidates nor companies can afford. In a tightening market, companies need to explore innovative talent pools to stay ahead of the competition – but too often bias gets in the way. The current hiring model is broken.

Enter artificial intelligence. AI can help create a level playing field for all candidates, regardless of their socioeconomic background. As Myra Norton, President and COO of Arena, which uses predictive analysis to place candidates into health care organisations, argues, “Talent is equally distributed, but opportunity is not, and data can help us navigate away from unconscious biases."

By analysing data-points to decide whether to put a candidate forward, artificial intelligence can ignore the inherent biases that a human recruiter may unknowingly, or knowingly, use to make the same decision. Case in point, recent studies have shown that in the US labour market, traditionally African-American names are systemically discriminated against, while ‘white’ sounding names are more likely to receive callbacks for interview.

A smiling black woman stands in front of a diverse group of professionalsJason Oxman, President and CEO of the Information Technology Industry Council, reasons that AI can also help eliminate biases facing candidates with criminal records. He believes, “when AI can help sift through a diverse pool of future employers in a fair way, I think we will see an enormous shift in the hiring process.”

AI may also be a blessing for hiring managers and recruiters. Despite the seconds spared for each one, collectively screening resumes is the most timing consuming aspect of the hiring process. 52% of talent acquisition leaders say the hardest part of recruitment is identifying the right candidates from a large applicant pool.

AI can reinforce biased hiring 

Like anything, AI has its limitations. Hiring managers should not rely on it alone to make decisions on who to take forward for an interview. Far from democratizing the workplace, an algorithm can reinforce biases if the data sets used feed on unfair or correlative human judgment-values, rather than objective and accurate metrics. If algorithms are built upon ‘successful’ applications that feature fewer resumes from women or minorities, the algorithm will replicate those biases for future applications and favour those from white, male applicants.

For example, the algorithm that Amazon employed between 2014 and 2017 to screen job applicants reportedly penalised words such as ‘women’ or the names of women-only colleges. Facebook’s housing and employment ads delivery have also been demonstrated to follow gender and race stereotypes, thanks to research conducted by Northeastern University, the University of California, and Upturn, a public-interest advocacy group.

A black man smiles and shakes hands with an Asian woman waiting for interview“What you have to look at are the measurements of success, and you have to make sure those don’t have biases incorporated,” says Jacob Hsu, CEO of Catalyte, “the core problem is trying to define excellence and then measure it in a fair way.”

Beyond the CV, hiring managers and recruiters are very good at communicating what’s needed for a role and extracting information from candidates during interview, but they can weigh the results poorly. In a meta-analysis of 17 studies of applicant evaluations, researchers Nathan Nathan Kuncel, Deniz Ones and David Klieger found that a simple equation outperformed human decisions by at least 25%. This applied to any situation with a large number of candidates, even for C-Suite positions. Interviewers, they said, are too often “distracted by things that might only be marginally relevant, and they use information inconsistently.”

Should you use AI to assist your hiring process?

Artificial intelligence holds mass potential for the recruitment process. If used correctly, it can supercharge and democratize hiring for all. As technology recruitment specialists, we are always working hard to keep you up-to-date with anything that may disrupt your talent acquisition process. Discover whether you should apply artificial intelligence to any stage of the hiring journey. Download our report, The future of AI recruitment, today.