Bias in algorithmic decisioning

Algorithmic bias is unintended prejudice in how an algorithm is designed, trained, or implemented. It can produce harmful outcomes, reinforce existing inequalities, and weaken fairness in data-driven decision-making.

Example

A hiring platform that uses an algorithm to rank job applicants might introduce bias if the training data reflects past discrimination. For example, if the dataset historically favored male applicants over female applicants, the algorithm may continue to rank male applicants higher. That carries gender bias into the hiring process. Reducing algorithmic bias is important for fairer outcomes in automated decision-making systems.