Algorithmic bias occurs when which of the following is true?

Algorithmic bias occurs when which of the following is true?

• A. A computer system reflects the implicit or explicit values of the people who collect, select, and use the data it generates, leading to discrimination
• B. It always eliminates bias
• C. It occurs only in theory
• D. It has no impact on decision making

Answer: (A)

Algorithmic bias occurs when a computer system encodes the values and assumptions of the people who collect, curate, and use the data and models, leading to discriminatory outcomes. The data used to train or tune these systems often reflect historical decisions, social inequalities, or biased labeling. As a result, the system may produce decisions or recommendations that systematically advantage or disadvantage certain groups, especially around race and other protected characteristics. In media practice this can show up in how content is prioritized, who gets visibility, or which audiences are served with certain messages, potentially reinforcing stereotypes or unequal access to information. Bias can creep in through the data itself, through design choices, through how success is defined, or through feedback loops that amplify disparities, so bias isn’t automatically removed by technology and can persist or worsen if not checked. This isn’t just a theoretical concern; it has real effects on decisions and opportunities people encounter. The idea that algorithms automatically eliminate bias is incorrect, and the notion that bias has no impact on decision making is equally false.

Algorithmic bias occurs when a computer system encodes the values and assumptions of the people who collect, curate, and use the data and models, leading to discriminatory outcomes. The data used to train or tune these systems often reflect historical decisions, social inequalities, or biased labeling. As a result, the system may produce decisions or recommendations that systematically advantage or disadvantage certain groups, especially around race and other protected characteristics. In media practice this can show up in how content is prioritized, who gets visibility, or which audiences are served with certain messages, potentially reinforcing stereotypes or unequal access to information. Bias can creep in through the data itself, through design choices, through how success is defined, or through feedback loops that amplify disparities, so bias isn’t automatically removed by technology and can persist or worsen if not checked.

This isn’t just a theoretical concern; it has real effects on decisions and opportunities people encounter. The idea that algorithms automatically eliminate bias is incorrect, and the notion that bias has no impact on decision making is equally false.