Unconscious bias has always been a challenge in the hiring process, with few solutions to overcome it. In recent years, Artificial intelligence (AI) has been touted as a tool that can help remove unintentional bias from hiring, but is it working? Some critics argue that because AI is “trained” by humans, humans’ bias creeps into the process, rendering the tools as errant as humans themselves.
AI Hiring Bias In Action
One of the most widely-publicized AI failures in hiring came from Amazon in 2018. The e-retail and tech behemoth had built a proprietary AI platform for recruiting and hiring that was touted as a means to remove the bias from hiring. At the time, the tech industry was under heavy scrutiny for lack of diversity in the industry, specifically gender diversity.
Amazon’s engine was built to scan through incoming resumes and observe past job postings candidates across a 10-year period. The problem was that most applicants during that time had been men, and the program began to penalize resumes that included gender-based keywords related to women’s clubs or women’s colleges.
In an even more embarrassing incident in 2016, Microsoft deployed an AI chatbot named Tay that interacted on Twitter. It was supposed to get smarter with each interaction, but users ended up teaching Tay profanity, inappropriate language, and race-based bias.
Is AI In Hiring Really All Bad?
Amazon and Microsoft are extreme examples of the AI conundrum. Whatever AI is taught, it learns. If there is already a skewed data pool like a candidate list of mostly men or if the AI bot is being taught to be downright racially biased, it will exacerbate the problem rather than solving it.
These examples show the industry that there is a distinct learning curve for AI, and there are critical points in the process where AI must be evaluated to improve outcomes. By improving the ways AI learns and interprets data, there are ways to eliminate bias.
Many companies are using AI forms that take simple actions like removing all age-related information from resumes or removing names from resumes to avoid gender bias or the phenomenon of rejecting people with names that “sound” like minority names.
AI can also help produce truly objective questions based on the job itself, rather than relying on human interviewers to create potentially biased questions. This helps generate quality data that can be evenly and objectively compared.
No matter what happens in the future, AI will likely never take the place of human recruiters and hiring managers. The technology will have to be continually tweaked and evolved to remove bias in the process but without creating its own unintentional bias. When deployed strategically, AI can enhance the hiring process and ensure hiring managers make the company’s right decision while organically cultivating a more diverse workforce.
Are You Looking To Improve Your Hiring Process For Your Sales Team?
If your industrial safety or PPE manufacturing company is looking for top-tier sales talent, talk to The Charis Group. As the industry’s premier recruiters, we help sales leaders build significant, diverse, and lasting teams. Contact us today to learn more.