A hiring algorithm rejects candidates based on zip code. Analysis reveals that in this city zip codes strongly correlate with race due to historical housing segregation even though race is not a feature in the model. Which type of AI bias does this scenario illustrate?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Think of it this way: in real-world AI governance, indirect or proxy bias is exactly what teams reach for when they need to handle this scenario. Indirect or proxy bias occurs when a seemingly neutral variable correlates with a protected attribute due to societal or historical factors. On the exam, remember that this falls squarely under the 1.0 AI Governance Overview domain.
Full explanation below image
Full Explanation
Indirect or proxy bias occurs when a seemingly neutral variable correlates with a protected attribute due to societal or historical factors. Using zip code as a proxy for race introduces discriminatory outcomes even though race is never explicitly referenced in the model. The correct answer, "Indirect or proxy bias", directly addresses the scenario described because it aligns with the specific governance requirement in question. The incorrect options ("Direct bias", "Training data imbalance", "Equalized odds") may seem plausible but do not satisfy the core requirement. Understanding the distinction between these concepts is critical for IBM watsonx.governance implementations and is frequently tested in the 1.0 AI Governance Overview section of the certification exam.