When the AI Is Biased, How Long Do Humans Look Before Agreeing?

Resume Screening, Fast and Slow: (Biased) AI Recommendations’ Influence on Human Decision Making

Kyra Wilson, Mattea Sim, Anna-Maria Gueorguieva, Soham Chatterjee, Aylin Caliskan · arXiv:2606.22213 (PDF) · submitted June 20, 2026

Summary by Andrew Johnson, OneTwo Resume editorial team · updated · We are not affiliated with the authors. arXiv papers are preprints and may not be peer-reviewed.

Most automated hiring ends with a human who is supposed to catch the machine’s mistakes. This 2026 study (from the same UW line as the embedding-bias audit) analyzes viewing-time data from an experiment on biased AI resume screening: how long people actually look at candidate resumes when an AI recommendation is present, and how that interacts with fairness of outcomes.

The core measured relationship: time spent viewing a resume corresponds to a candidate’s selection chances — attention is the currency, and the study examines how AI recommendations redistribute it. It’s the empirical middle ground between "human in the loop fixes everything" and "humans just rubber-stamp the machine."

What the paper reports

What this means for your resume

Our editorial interpretation — the paper does not give job-seeker advice.

Read it with these caveats

arXiv preprint (v1, June 2026); experimental setting with study participants rather than professional recruiters at work — attention dynamics may differ under real hiring incentives.

Primary source: Resume Screening, Fast and Slow: (Biased) AI Recommendations’ Influence on Human Decision Making — always read the paper before citing it. Spotted an error in our summary? Tell us and we'll fix it with a visible correction.

More in humans + ai deciding together