Several organizations are struggling to discover strategies for recruiting a more diverse workforce, and a few are turning to artificial intelligence (AI). But artificial intelligence obtained a bad rap past year when news got out that Amazon’s internal AI recruiting tool had”learned” sex bias. So, is AI beneficial to those looking for diversity, or will it only exacerbate the problem? One recruitment firm has discovered that AI is an effective strategy for increasing the diversity of candidate pools, provided that it’s implemented correctly.
The Problems With Using AI In Recruiting
Amazon scrapped its internal AI recruiting efforts after realizing the test recruitment program that they designed was biased against women. The computer models were trained using résumés filed to Amazon over the previous ten years. Unsurprisingly, the majority of these résumés came out of men. Therefore, the computer versions”heard” that men were superior job candidates.
Because of this, the models became more biased against girls, generating lower evaluations to graduates of two all-women’s schools and penalizing résumés that included the term”women” as in”women’s club captain” Even though the applications could be changed to become impartial to these particular terms, Amazon decided to scrap the job out of fear that the models may create other discriminatory selection criteria.
The Benefits Of Using AI In Recruiting
Despite Amazon’s setbacks, Genevieve Jurvetson, founder and CMO of the artificial intelligence-based recruitment company, Fetcher, considers AI is a vital tool for those looking for a more diverse candidate pool. “If you get too involved in the fears surrounding AI, you overlook a huge opportunity,” Jurvetson says. She explains that there are innumerable criteria that you may use to find good candidates that aren’t remotely gender-related. Jurvetson asserts that Fetcher’s usage of AI techniques permits them to bring minorities and women in their clients’ hiring pipeline in a way other recruiting methods just can’t.
Why are AI techniques so effective? Most of us know people have biases. Take sex bias, for example. Following a lifetime of exposure to largely guys in leadership roles, in boardrooms and in technology jobs, we develop a bias to prefer men in such places. The large benefit of artificial intelligence is that it eliminates the people and their biases out of a large chunk of the recruiting process. The more you can remove human intervention, especially at the stages of the recruiting process that are most likely to prejudice, the less that bias will be able to influence decision-making.
Though Fetcher permits clients to”thumbs up or thumbs down” potential applicants, Jurvetson believes the most diverse candidate pools come in what they call their fully automatic mode. In this mode, Fetcher hones on customer preferences by asking the customer to review a small number of candidates. Afterward, Fetcher’s systems require control, building a candidate pool, and contacting potential candidates directly. In this manner, a client’s possible biases are removed from the method of choosing who will be contacted. Jurvetson explains,”When you pull yourself from this procedure, that’s a really important step because if you’re not hand-selecting each candidate, you’re not earning these inherent biases which most of us have.”
One reason that associations have trouble increasing diversity is that the standard procedures of looking for candidates are often biased. The problem, Jurvetson describes, is that searches often use bias-ridden proxies for offender potential. “Can they go into a top 20 college? Did they come out of a top-tier firm? Those are normally pools of talent which may not be that varied to start with. Using AI to be able to spot patterns, career progression is a great one, that correlate with success better than those previous proxies is effective and exciting,” she clarifies.
In accordance with Jurvetson, searches utilizing these less-traditional proxies for achievement characterized by artificial intelligence applications allow Fetcher to locate candidates which may otherwise go overlooked. This enables organizations to throw a broader net than they could otherwise.
Research confirms there are huge advantages of casting a wide net when assembling a candidate pool if you would like to increase diversity. One study discovered that if you only have one female candidate or underrepresented minority candidate on your pool of candidates, they have practically no chance of making it to the offer stage. That is because the only girl or minority appears too different from the norm. But if you include a second female or minority candidate, their probability of making it to the final round increase dramatically, and they have the same chance of getting an offer as the other candidates.
What retains the Fetcher models from having the exact same difficulty as Amazon’s unsuccessful model? Jurvetson suggests that precautions will need to be taken to minimize any unintended effects. At Fetcher, she says they keep their versions easy, and they keep a trained human eye on the models and the outputs to be sure the recruiting criteria are gender-neutral and the models are coming diverse candidate pools.
The Future Of AI In Recruiting
Recently, Tomas Chamorro-Premuzic and Reece Akhtar created a disagreement in Harvard Business Review for carrying the use of AI techniques one step farther and implementing them into the interview procedure. The investigators report,”One of the major problems with the way we now interview job applicants is that the procedure is largely unstructured, leaving the coughing to the whims and fancies of their interviewer. It shouldn’t take much convincing to realize how this isn’t just inefficient, but it also leads to biased conclusion as a result of interviewers expressing and seeking to confirm their own tastes. This is the area where digital or video interviews are likely to help. Digital interviews can get rid of these constraints almost entirely.”
While optimistic at about their potential, Jurvetson does not think AI occupation interviews will take over any time soon. “I think if used thoughtfully, they may be really powerful and beneficial for a whole lot of candidates, especially at early stages of the recruiting process, however I don’t have to inform you all of the ways that may go wrong also. You have to be smart about how and when you utilize these kinds of interviews, or racial or gender issues can come into play in a really sad way, and you might also negatively impact the candidate experience,” she reports. Yet another limitation she added,”I’ve heard from candidates interviewing for more experienced jobs that they sometimes find [automatic interviews] insulting, since they can give the impression that the potential employer is not eager to provide you their time.”
For those looking to produce a more diverse workforce, eliminating human decision-making from at least a component of the recruitment process seems like a no-brainer for reducing unconscious bias in hiring. Implemented correctly, AI tools can be a great way to search in an unbiased way. And for those job seekers of the future, keep an open mind about a robot interviewer–ideally it will be programmed to be more biased than the human manager.