Definition
AI bias and fairness encompasses systematic disparities in how AI systems represent, evaluate, or treat demographic groups — including gender, race, age, and other protected characteristics.
Age Dimension (2026)
Prior research focused heavily on gender and racial bias. kaist’s June 2026 study is among the first quantitative analyses of age-related stereotypes in conversational AI (digital-ageism).
Stereotype Content Model dimensions:
- Warmth: sociability, morality, kindness
- Competence: ability, assertiveness, expertise
Mitigation Approaches
- Diverse generational participation in AI development (Choi, KAIST)
- Bias evaluation across demographic dimensions beyond gender/race
- Responsible AI auditing for production LLM deployments