Examples
Worked examples
- Is an instance
An LLM that consistently misgenders historical scientists from non-English-speaking countries
- Is an instance
An image classifier with 30% lower accuracy on darker-skinned subjects
Counter-examples
Looks similar, but isn't
- Not an instance
A model that is uniformly wrong across all groups exhibits inaccuracy but not bias in this sense
Editorial commentary
For scholarly disclosure, the relevant question is whether bias in the AI tool used could have materially affected the work’s findings or claims. Examples include LLMs underrepresenting non-Western scholarly traditions in literature summaries, or vision models with poorer accuracy on darker skin tones.
References
- Mehrabi et al. 2021 ‘A Survey on Bias and Fairness in Machine Learning’ ACM Computing Surveys
- Buolamwini and Gebru 2018 ‘Gender Shades’ FAT*
Also known as
Algorithmic bias · Model bias
Machine-readable encodings
Use in your systems
xml
<role vocab="credit"
vocab-identifier="https://casrai.org/dictionary/"
vocab-term="AI bias"
vocab-term-identifier="https://casrai.org/dictionary/term/ai-bias" />json
{
"@context": "https://schema.org",
"@type": "DefinedTerm",
"name": "AI bias",
"identifier": "https://casrai.org/dictionary/term/ai-bias",
"description": "Systematic skew in an AI system's outputs that produces unjustified differential treatment, accuracy, or representation across groups, tasks, or contexts, arising from training-data composition, model architecture, objective function, or deployment context.",
"inDefinedTermSet": "https://casrai.org/dictionary/domain/generative-ai-use-and-disclosure/",
"url": "https://casrai.org/dictionary/term/ai-bias",
"sameAs": [
"Algorithmic bias",
"Model bias"
],
"license": "https://creativecommons.org/licenses/by/4.0/"
}
