AAIA Question 27

Community Discussion

Written by other learners. Comments can disagree with the marked answer — check the official documentation when in doubt.

• Chose C1

  Data governance is the foundational process that ensures the inputs are high-quality, which is the only way to ensure the outputs are credible. Trustworthiness starts at the source: For an AI output to be credible, the training data must be accurate, complete, representative, and timely. Data governance provides the framework for data quality, integrity, and lineage. Auditability and Verification: Governance provides the "paper trail." If an AI model produces a questionable result, an organization with strong data governance can trace that result back to the specific data points that influenced it. Bias Mitigation: Credibility is often destroyed by hidden biases. Data governance includes processes for identifying and cleaning biased datasets before they ever reach the model.

• Chose B1

  The answer is not D, but B for the following reasons. Please delete my original answer, as it has not yet been approved. The process by which a model is adjusted to produce output that is consistent with the organization's objectives and rules. → This directly ensures reliability.

• Chose D1

  Data analysis verifies the following aspects of AI output: Accuracy Consistency Outlier detection Model performance evaluation Bias detection Output validity testing In other words, this is the process of directly verifying whether AI output is trustworthy. To ensure the reliability of output, data analysis is the most practically necessary process.