It is recommended to carefully consider which demographic data and for what purpose is fed...

The correct answer is A. To avoid societal bias. The uploaded study guide directly identifies societal bias as the concern when demographic data is selected for AI-model use. Demographic information can contain or reflect historical social inequalities. When characteristics such as age, gender, location, ethnicity, socioeconomic indicators, or related proxies are included in a model without careful justification, the model may learn patterns that reproduce existing disparities rather than simply represent neutral relationships. The UX and responsible-design implication is that teams must question both what data is being collected and why the model needs it. Data should not be introduced simply because it is available. Instead, each attribute should have a legitimate purpose that supports the intended user or business outcome without unnecessarily increasing the risk of discriminatory behavior. Option B refers to confirmation bias, which concerns a tendency to favor evidence that supports existing beliefs. While that can affect research and decision-making, it is not the specific data-governance risk described here. Option C, interaction bias, relates more to patterns arising from how people interact with a system rather than demographic information itself. The study-guide principle is therefore that responsible AI design requires deliberate demographic-data selection to reduce the propagation of societal bias into automated decisions.

Study Guide reference/topics: Responsible AI, demographic data, societal bias, responsible design, Human-Centered Design.

==================================