Episode Details

Every AI hiring vendor promises the same safeguard: a human reviews the AI's recommendation before any candidate advances or gets cut. It sounds responsible. But peer-reviewed research from 2023 to 2025 is telling a very different story. In one study, 528 participants evaluated hiring decisions alongside a deliberately biased AI — and followed its picks roughly 90% of the time. That's not a human-in-the-loop. That's a rubber stamp. And it turns out, the faster and busier your recruiting team is, the worse the problem gets. In this episode, we dig into the science of automation bias in hiring: what it is, why time pressure dramatically amplifies it, and what structural fixes the research says actually work. Plus — what regulators including the EEOC, the EU AI Act, and NYC Local Law 144 are now requiring beyond the checkbox.