Several years ago, AI started to be hailed as a new way to screen candidates in recruiting, enabling a more effective way to sift through high volumes of job applications to find the most qualified candidates. However, we have since learned that AI is only as unbiased as its programmers, and bias is showing up in big enough ways to enact new legal protection.
New York City just passed a law that will “ban employers from using automated employment decision tools to screen job candidates, unless the technology has been subject to a ‘bias audit’ conducted a year before the use of the tool… Companies also will be required to notify employees or candidates if the tool was used to make job decisions,” (bloomberglaw.com). Illinois, Maryland and Washington, D.C. have also introduced similar legislation.
This is a big win for people of color and women, who have been shown to be discriminated against by many AI-based screening tools.
Amazon acknowledged it scrapped its AI recruiting tool when it was shown to discriminate against women. Many lawsuits have been filed with the Equal Employment Opportunity Commission (EEOC) against other companies with similar charges based on race, gender and age.
These issues show yet another way unconscious bias can create highly detrimental impacts. I do believe there is a way for these tools to serve employers and potential employees well. They need to lead with design for intentionally stripping out bias, and then solve for the need for efficiency. And diverse design teams are a core requirement to ensure that happens.
Michelle’s mission is to help companies create equitable workplaces. She is the Founder and CEO of Equity At Work, helping leaders achieve major impact through their diversity, inclusion and equity work. Follow her on LinkedIn and Twitter.
Photo credit: Eric Prouzet