CyberCertPrepAI Security FundamentalsAI & ML Concepts for SecurityEASYFree question
A security researcher wants to understand whether an AI system's decisions could be challenged legally or ethically. Which concept describes the degree to which an AI model's decisions can be understood and explained?
- A.A. Explainability (XAI) - the extent to which the internal logic of an AI decision can be understood by humans
- B.B. Interoperability - the ability of the AI model to integrate with other systems
- C.C. Scalability - the AI system's ability to handle increasing workloads
- D.D. Portability - the ability to run the model on different hardware platforms
Why A is correct
Explainability (or Interpretability, collectively XAI - Explainable AI) refers to how well the reasoning behind a model's decision can be understood and communicated to humans. This is critical for legal challenges (e.g., GDPR Article 22 right to explanation for automated decisions), bias audits, and building stakeholder trust. The EU AI Act mandates transparency and explainability for high-risk AI systems. Interoperability (B), scalability (C), and portability (D) are system quality attributes unrelated to decision explainability.
Know someone studying for AI Security Fundamentals? Send them this one.