AI Has Broken Hiring. Here’s How to Fix It HBR Article 2026
The Harvard Business Review article by Shraddha Sunil and Mudit Saraf (June 8 2026) documents how generative AI is eroding the reliability of classic hiring signals—polished résumés and remote first‑round interviews—by enabling candidates to fabricate or augment their presentations in real time. The authors conducted a mixed‑method operational study: 120 interviews with talent‑acquisition leaders across 87 companies and analysis of 6,380 recorded first‑round screening sessions. Findings reveal that AI‑generated résumés are biased toward the style of the screening model, inflating short‑list rates by 23‑60 %, and that real‑time interview assistance tools (e.g., Final Round AI, Parakeet, Cluely) produce suspicious patterns—latency anomalies, vocabulary shifts, and gaze inconsistencies—in up to 60 % of early‑career technical interviews. The breakdown of early‑stage filters drives recruiters toward closed networks, harming diversity, and raises strategic risk for the C‑suite as talent density declines and hiring costs rise. To counteract these threats, the authors propose five shifts: (1) separate authenticity checks from skills assessment; (2) replace static behavioral screens with adaptive, reasoning‑based first‑round interviews; (3) allow AI tool use in later‑stage interviews to evaluate judgment and hallucination detection; (4) foster a culture of intellectual honesty over polished perfection; and (5) retain in‑person final rounds as targeted audits. The article also notes emerging practices such as Meta’s AI‑enabled coding interviews, which aim to test candidates’ ability to work with AI rather than without it.
Topics
AI‑Generated Résumés Undermine First‑Round Signal Value
Explains how generative AI lets candidates produce keyword‑rich, stylistically aligned résumés in minutes, and how screening models preferentially rank résumés that match their own output style.