Artificial Intelligence
Bias (AI)
Also known as: Bias in AI & Algorithmic bias
Plain English
When AI is unfair or skewed.
Definition
glossaryBiasBias is a systematic distortion in thinking or data that affects the accuracy of research or decision-making.Open glossary term in AI refers to systematic errors or unfair outcomes caused by skewed glossaryDataData is raw, uninterpreted information collected and stored so it can be analysed, processed, or used to inform decisions. On its own it carries no meaning; context and interpretation are what make it useful.Open glossary term, assumptions, or glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term design.
In practice
Identified through testing outputs across different scenarios and user groups.
In context
It is rarely introduced deliberately. It arrives through a glossaryDatasetA dataset is a structured collection of data used for analysis, training models, or processing.Open glossary term that reflects historic decisions, and the glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term reproduces those decisions with more glossaryConsistencyConsistency is the use of uniform patterns, behaviours, and visual elements across a product to create familiarity and predictability. It helps users learn once and apply that knowledge throughout the experience.Open glossary term than the humans ever managed.
The reality
glossaryBiasBias is a systematic distortion in thinking or data that affects the accuracy of research or decision-making.Open glossary term is difficult to eliminate completely and often reflects underlying issues in guideTraining DataWhat training data is, how it shapes what an AI model knows and assumes, and what product and design teams need to understand about its role in quality.Open guide.
Compared with
Bias in AI vs Hallucination
glossaryBiasBias is a systematic distortion in thinking or data that affects the accuracy of research or decision-making.Open glossary term is a systematic skew, consistently wrong in a particular direction. A hallucination is a fabricated output. Bias is a glossaryPatternA pattern is a reusable solution to a design problem that recurs across products and contexts. It captures an approach already shown to work, so a team solves the problem once rather than every time it appears.Open glossary term you can measure; hallucination is an event you have to catch.
FAQ
Common questions
A few practical answers to the questions that usually come up around this term.
What is bias in AI?
It is when AI produces unfair or skewed results due to glossaryDataData is raw, uninterpreted information collected and stored so it can be analysed, processed, or used to inform decisions. On its own it carries no meaning; context and interpretation are what make it useful.Open glossary term or design.
What causes AI bias?
Biased guideTraining DataWhat training data is, how it shapes what an AI model knows and assumes, and what product and design teams need to understand about its role in quality.Open guide or flawed assumptions.
Why is AI bias a problem?
It can glossaryLeadA lead is a potential customer who has shown interest in a product or service, typically by providing contact information or engaging with content.Open glossary term to unfair or inaccurate outcomes.
How can AI bias be reduced?
By improving glossaryDataData is raw, uninterpreted information collected and stored so it can be analysed, processed, or used to inform decisions. On its own it carries no meaning; context and interpretation are what make it useful.Open glossary term diversity and testing outputs.
Related Services
Related Terms
