The MAESTRO model in AI Threat Modeling emphasizes what?

• A. Data normalization and schema design
• B. AI-specific categories and adversarial machine learning
• C. Marketing risk and brand protection
• D. Hardware reliability and fault tolerance

Answer: (B)

The main idea being tested is recognizing that threat modeling for AI systems targets risks unique to AI, not just general IT risks. MAESTRO centers on AI-specific threat categories and adversarial machine learning, focusing on how data and models can be manipulated to produce harmful outcomes. This includes threats like tampering with training data, manipulating inputs to cause incorrect model behavior, and attempts to steal or extract knowledge from the model. Because these adversarial and data-centric concerns are unique to AI systems, they’re the core focus of MAESTRO, making the option that highlights AI-specific categories and adversarial machine learning the best fit. The other choices describe broader concerns—data schema design, marketing/brand risk, and hardware reliability—that aren’t about AI threat vectors.