AI · 18h ago
AI hiring tools can develop their own biases beyond human data
New research shows LLMs used in hiring can form biases independent of training data, not just mirroring human prejudices. These AI-specific biases may lead to unfair candidate screening even when human biases are removed. The findings challenge assumptions that AI hiring tools are inherently less biased than human recruiters.
Meridian48 take
The study suggests AI bias isn't just a data problem—it's an emergent behavior that could make automated hiring even harder to regulate than previously thought.
Read the full reporting
AI is more likely than humans to form biases when hiring →
MIT Technology Review
ai-biashiring-algorithms