Which phrase refers to systematic errors in AI that produce unfair results due to biased data or design?

Which phrase refers to systematic errors in AI that produce unfair results due to biased data or design?

• A. Bias in AI
• B. Fairness in AI
• C. Transparency in AI
• D. Privacy in AI

Answer: (A)

Systematic unfairness in AI due to biased data or design is bias in AI. When training data reflect historical prejudices or underrepresent certain groups, or when the model’s design and objective choices encode those biases, the model learns patterns that lead to unequal or discriminatory outcomes. This causes predictions or decisions to favor some groups over others, often in ways that aren’t warranted by the task. Fairness in AI is the goal of achieving equitable outcomes and involves techniques to reduce bias, while transparency focuses on making how models work understandable, and privacy concerns protect data from misuse. So the phenomenon described is bias in AI.

Systematic unfairness in AI due to biased data or design is bias in AI. When training data reflect historical prejudices or underrepresent certain groups, or when the model’s design and objective choices encode those biases, the model learns patterns that lead to unequal or discriminatory outcomes. This causes predictions or decisions to favor some groups over others, often in ways that aren’t warranted by the task. Fairness in AI is the goal of achieving equitable outcomes and involves techniques to reduce bias, while transparency focuses on making how models work understandable, and privacy concerns protect data from misuse. So the phenomenon described is bias in AI.