Algorithmic monoculture can cause systemic rejection.
We also explore new concerns posed by common reliance on a single hiring vendor. in our before workWe theorized that an algorithmic monoculture, where many employers come to rely on the same algorithmic recommendations, could lock some people out of jobs. We use a large dataset of real-world recruitment AI recommendations to test our hypotheses. Those who have submitted more than one application for a selected position may same Algorithmic hiring vendors are more likely to be rejected from every position they apply to than if each company made their decisions statistically independent of each other. Ten percent of applicants who submit four applications are rejected from every place they apply.
Our research also found that this pattern did not hold true in other situations. We analyzed the largest amount of data ever study The study, which received 83,000 applications across 108 Fortune 500 companies during the same period as our study, did not focus on whether AI was used in decision-making. This data shows that the rejection rate for applicants across all the companies to which they applied is no higher than would be expected if each company made its decisions independently of the others.
This suggests that market concentration is important. When a single recruiting vendor dominates industry selection, candidates are more likely to be squeezed out.
