Episode Details

A new Stanford-led study just dropped — and it's the largest analysis of AI hiring algorithms ever conducted. Four million job applications. 156 major employers. One vendor. And the finding that should stop every HR leader cold: one in four Black applicants is being screened by an AI system that produces racially biased outcomes. The researchers didn't just find bias — they found that the way most companies audit for it is fundamentally broken. Aggregate audits hide exactly the kind of per-position discrimination this study uncovered. And there's a second twist: something called the "algorithmic blackball" effect, where a single low score follows a candidate across dozens of employers for up to a year. In this episode, we break down what the study actually found, why the legal exposure is suddenly very real — with the EU AI Act enforcement deadline hitting August 2, 2026 — and five concrete things HR teams should do right now to protect their organizations and their candidates. This one is important. Share it with your legal and recruiting leads.