Algorithmic Fairness

Algorithmic fairness encompasses a set of mathematical and ethical frameworks for ensuring AI systems do not unfairly advantage or disadvantage particular groups. Key fairness metrics include demographic parity (equal selection rates across groups), equalized odds (equal true positive and false positive rates), and calibration (equal predictive accuracy across groups). In AI hiring, these concepts translate to ensuring that candidate screening, scoring, and ranking tools produce equitable outcomes across race, gender, age, and other protected characteristics. Different fairness metrics can conflict, requiring context-dependent choices about which metrics to prioritize.