AI In Hiring Processes: The Biases and Solutions

World Economic Forum

Artificial Intelligence (AI) is becoming increasingly prevalent in our every day lives, and has recently found its way into hiring practices all across the United States. In hopes of finding potential employees more efficiently, many employers are using machine-learning algorithms to search through social profiles and online resumes to find the best person for the job based on set characteristics. As objective as this algorithm may seem, it can play a very negative role.

On the surface, the use of AI in hiring practices seems nothing but helpful. After all, what’s wrong with saving some time and money? However, studies have shown that this new technology raises several ethical issues, particularity by exacerbating unjust gender and racial biases.

When it comes to applying for positions at tech companies like Amazon, women experience a fair disadvantage when it comes to applicant selection using AI. Amazon’s “AI recruitment engine” was found to direct bias against women’s applications. The engine would go through applications and penalize resumes that included the word “women’s,” as in “women’s chess club captain.” Additionally, it penalized graduates of all-women colleges, according to Reuters.

The selection of job applications using AI systems like the Amazon AI recruiter engine further promote gender bias in the workforce and exclude women from certain fields. In Big Tech companies like Microsoft and Google, women make up only a fraction of total employment, and AI involvement is only making this gender disparity worse.

Additionally, AI’s use in hiring practices may not be as efficient and helpful for applicants as it is for employers. When an applicant’s resume is processed through a tracking system or engine, AI is used to determine keywords within the resume. If the resume has words that are not search engine optimized, the resume will most likely be rejected. This selection can negatively impact job seekers if they are applying and unaware of the use of AI in their application.

A wide variety of solutions have been proposed by companies and legislators alike to counteract the negative effects of AI bias. Some companies have implemented programmed “nudges,” which are specifically implemented to remind a program that it should be considering certain groups equally. This works to correct bias, but still does not fully account for the flaws of such software, and these practices are not adopted by the majority of companies utilizing AI in hiring practices.

Illinois has been quite progressive in encouraging transparency surrounding hiring practices. Legislation “requires companies to inform applicants that AI will be used to analyze their interview videos.” (Waltz, 2019) However, it isn’t clear whether this actually protects the applicants from the bias AI has been proven to exhibit. (Jimenez, 2020) Laws also protects applicants from immediate disqualification if they don’t want their applications subject to AI examination. This legislation should absolutely be the model in order to move toward a future with more transparency and concern for the wellbeing of people affected by AI’s biases.

As dystopian as it seems, AI and machine learning is advancing our world, however leaving unerasable marks on people’s lives changing their lives many times for the worse. We must hope to make change with AI so that our globe continues to spin in the direction of prosperity and success.