AI Glossary

AI bias

Also called algorithmic bias

What does “AI bias” mean?

AI bias: Systematic skew in an AI system's outputs that unfairly favours or disadvantages certain groups, viewpoints or kinds of input. It usually comes from patterns in training data, choices about what to measure, or how a system is deployed, rather than from any deliberate intent.

AI bias visual guide: Systematic skew in an AI system's outputs that unfairly favours or disadvantages certain groups, viewpoints or kinds of input. It usually comes from patterns in training data, choices about what to measure, or how a system is deployed, rather than from any deliberate intent.
Visual guide Open full size (new tab)

What is an example of “AI bias”?

An image generator that mostly depicts doctors as men, because its training images did, shows bias inherited from data.

Keep learning

See also

FairnessTermResponsible AITermData poisoningTermModel cardTerm

Same topic

LLM-as-a-judgeTermAI regulationTermAI safetyTermAlignmentTermArena rankingsTerm