The term AI bias belongs to the fields of artificial intelligence, big data, smart data and digital society. AI bias means that artificial intelligence - i.e. learning computer systems - develops certain prejudices and makes incorrect decisions because it has been fed incorrect or one-sided data.
The problem usually arises when the data sets used to train the AI are not balanced. Let's assume a company uses an AI to screen job applications. If the AI has mainly received information from previous, predominantly male employees during training, it could unconsciously put female applicants at a disadvantage. This is not because the AI itself is malicious, but because it draws the wrong conclusions from the unevenly distributed examples.
AI bias is therefore a real stumbling block when it comes to achieving fair and objective decisions through artificial intelligence. Companies should therefore take care to select their data carefully and regularly check their AI for bias so that no one is disadvantaged.
Deutsch 
English (UK)