AI’s potential to infect the hiring process with bias
She’s also seen where AI might use historical data to determine patterns of a successful employee, identifying certain schools, work histories, tenure, or other characteristics that, while not inherently bias, perpetuates the bias that accurate correlations exist between these elements, when they often don’t. Organizations need to ensure that humans remain a part of these processes, Cure adds, where they can bring context, intuition, nuance, and an ability to identify potential in a candidate that AI can’t replicate.
“I think we’re allowing AI to become the process instead of allowing it to support the process in ways that makes hiring better,” she says.
An emphasis on accuracy over speed
Cure says a major problem for most companies is that the balance is off, with companies using AI for the majority, if not all, of résumé screening rather than as a complement to human efforts. Organizations that simply implement AI to speed up different parts of the hiring process, without taking time to consider if a process stands to benefit from AI, run the risk of introducing bias.
