Glossary

Bias

What is Bias?

In AI, bias refers to a systematic distortion of results. It often stems from unbalanced or historically skewed training data, but also from model design or how a question is framed. A well-known example: an applicant screening tool that learns from past hiring decisions can perpetuate existing disadvantages. In statistics, bias also means the deviation of an estimate from the true value. Countermeasures include diverse data, regular audits and human decisions on personnel matters.

Topics

AI ethics
Fairness
Data quality

Try KI.ordo

Agents, flows and templates in one workspace – free 24-hour demo.

Start demo

← All terms

Related terms

Discrimination in AI

Discrimination in AI occurs when an AI system unjustifiably disadvantages people on the basis of characteristics such as gender, ethnic origin, age or disability. Causes are usually skewed training data, unsuitable proxy features such as postcodes, or a lack of testing; well-known cases come from recruitment and lending. Countermeasures include representative data, regular fairness testing, transparency and human oversight of decisions about people.

fairness
bias
+1

Training Data

Training data are the examples from which an AI model learns patterns: large text collections for language models, labelled images for image models. Their quality, diversity and recency largely determine how reliable and fair a model is and how up to date its knowledge will be. For organisations it also matters whether a provider uses the customer data they enter for further training.

Data
Training
+1

AI Ethics

AI ethics deals with the moral questions surrounding the development and use of AI: fairness and discrimination, transparency and explainability, data protection, responsibility for errors, safety and the effects on work and society. For organisations it becomes concrete in guidelines, risk assessments, human oversight and honest communication about AI's limitations. Example: before a CV screening tool is introduced, it is checked for whether it disadvantages particular groups.

Ethics
Responsibility
+1

Explainability

Explainability (explainable AI, XAI) describes how well one can understand why an AI system reached a particular result. It covers methods that show which inputs influenced a decision most, as well as practical aids such as source references and logs. Explainability matters for trust, troubleshooting and regulation – for example when a bank has to justify why an application was declined.

XAI
Transparency
+1