Basic Concept: Responsible AI is a governance framework addressing risks that arise from AI systems producing outcomes that are unfair, harmful, or contrary to human values. Different risk types fall under different governance domains — some under responsible AI, others under security or operational management. CompTIA SecAI+ Study Guide covers responsible AI risk categories under Domain 4.

Why C is Correct: Response bias occurs when an AI system ' s outputs are systematically skewed against certain groups, topics, or perspectives, reflecting biases embedded in training data or model design. This is a core risk addressed by responsible AI principles including fairness, non-discrimination, and explainability. Responsible AI frameworks mandate bias detection, assessment, and mitigation to ensure AI responses treat all users and groups equitably.

Why A is Wrong: Model drift describes the degradation of model performance over time as the distribution of real-world data diverges from the training data distribution. While an important operational concern, model drift is primarily a technical performance risk managed through MLOps and monitoring practices, not a core responsible AI governance concern.

Why B is Wrong: Reputational loss is a business risk consequence that may result from various AI failures including biased outputs or privacy violations. It is an outcome or impact rather than a specific risk category that responsible AI frameworks directly address.

Why D is Wrong: Data poisoning is a security attack where adversaries corrupt AI training data to manipulate model behavior. This is a cybersecurity threat managed through security controls and data integrity protections rather than responsible AI ethical governance frameworks focused on fairness and accountability.