Removing Unconscious Hiring Bias Using AI (INFOGRAPHIC)
Brian Wallace, Founder & President, NowSourcing
Many people consider artificial intelligence or AI to be our automated future. We have heard that it will take over some jobs but will ultimately facilitate our lives and make it easier. However, this might not be the case when it comes to increasing diversity.
Diversity is one of the most vital aspects of business. Only diverse businesses will grow and thrive in the future. Diversity brings revenue, increases innovation, and adds positive value creation to a business. Yet, businesses do not embrace diversity.
Both gender and racial biases exist in the recruitment and hiring process which is why some firms are attempting to switch to AI enhanced recruiting processes. However, AI recruitment brings a whole other set of biases into the recruiting process.
Bias such as Boolean search bias and resume building biases are created by software to artificially boost candidates who’s application and resume contain key phrases that the software is coded to look for. Applicants who don’t meet these requirements are commonly excluded.
Other biases result from human error. Coding AI software requires a lot of skill, less skilled workers could unintentionally make biased search questions based on their unconscious human bias.
AI however, can greatly reduce bias and increase diversity if utilized by companies effectively. In 2018, Amazon completely rewrote their AI recruiting software after it found significant bias against women in the code. Actions like this must continue to be done and lauded so that we can develop an effective bias-free software.
In order to create unbiased AI we have to have unbiased data. This means that we must change our sampling behaviors and methods to provide bias free data for our software. This could mean expanding the sample size, increasing sample diversity or location, and broadening the industry filters on recruiting data.
Factors like age, gender, and race should be excluded in the AI data sets. This can ensure that the software is making decisions based purely on skill sets and job effectiveness rather than physical traits.
With that being said, we need to have more categorical data such as interests, skills, and traits such as responsibility factored into the recruiting process. This needs to start even before the data is collective and involves a shift in the business resume writing mindset.
Finally, unbiased AI should be able to consider fit in a two-way perspective. It should be able to evaluate how a person’s skills will benefit and assist the position or company but, it should also be able to evaluate how the company will fit the person. Things like morals or values of both the person and the company need to be factored into the recruitment process in order to protect both sides and allow them to prosper.
No one can deny that the business world desperately needs to see some diversity in the field. Without it, many businesses simply will not survive the next ten years. Diversity means new ideals and new perspectives, which is something that could benefit the executive field tenfold.
About the Author: Brian Wallace is the Founder and President of NowSourcing, an industry leading infographic design agency in Louisville, KY and Cincinnati, OH which works with companies ranging from startups to Fortune 500s. Brian runs #LinkedInLocal events, hosts the Next Action Podcast, and has been named a Google Small Business Adviser for 2016-present. Follow Brian Wallace on Linked
