After affirmative action ruling, what role can AI play in hiring?

On June 29, the Supreme Court's decision on affirmative action ended race-conscious admissions practices in higher education. As employers scramble to anticipate what it could mean for the workplace, many are considering technology as a way to build and maintain diverse workforces. 

Sixty-five percent of recruiters currently use artificial intelligence in the recruitment process, according to workplace insights platform Zippia, and while 48% of hiring managers admit to having bias, 68% think that AI could be the key to removing that unconscious bias from the hiring process.

"If you went to a top school or had the best internships, that usually means you've had more resources," says Will Rose, chief technology officer at Talent Select AI. "If employers are basing hiring decisions on these traditional markers, it's not always helpful when trying to prioritize diversity. AI is able to bring more insights and data to the hiring manager to make a more holistic decision about a candidate as a whole." 

Read more: How the Supreme Court's decision on affirmative action will impact employers

Often, resumes and cover letters don't capture an applicant's personality or soft skills that could make them a good fit beyond their transcripts and achievements. Talent Select AI partners with workplaces and colleges to provide them with accurate, unbiased insights on potential candidates, using AI to analyze their interviews and pick up on language used by the applicant in their meeting. From there, a personality and competency profile is built for employers, which they can use to test a candidate's compatibility with their organization. 

"We're providing insights about strengths and competencies and skills that are predictors of job performance," Rose says. "We're aware that these candidates might be a little short on the traditional markers, but here's a lot more information on why they might be very strong in the traits you're looking for, and why this might be a great opportunity to make a hire that you would otherwise not make." 

Despite the many ways AI can assist in diversifying a company's ranks, there is still a very real threat that it could do the opposite. In 2022, USC researchers found bias in up to 38.6% of "facts" used by AI, such as associating women with negative qualifiers and gendered terms, and associating certain religions and races with discriminatory stereotypes. That kind of lapse in judgment from bots could result in unethical hiring practices that end up discriminating against applicants of certain genders, ethnicities, sexualities and religions. Should that be the case, it would not only worsen the aftermath of the affirmative action ruling, but could get employers in legal trouble with the Equal Employment Opportunity Commission. 

Read more: Using AI to recruit? You're legally responsible for the bot's bias, EEOC says

"There's always a potential risk of using any kind of machine-learning-based system if you leave it unchecked," Rose says. "But the great thing about these technologies is that there is a way to make sure that you're doing things the right way, and it's by not only making sure that you're not introducing biased data into the system, but that you're also accounting for it." 

Simply trusting that human beings can successfully keep the information they feed AI tools bias-free is a losing strategy, according to Rose. In order for AI to truly be able to combat and — hopefully — reverse bias in recruiting, there have to be continuous assessments manually and meticulously made against every subgroup the AI could potentially exclude using. Used correctly, AI tools can use data to source candidates objectively and according to diversity ethics and guidelines specifically set by companies.

"There's a well established history with assessments in the hiring space that has laid a good framework on how to successfully hire," Rose says. "Technology can make sure that framework and those measurements are done fairly and in the right way."