In AI Threat Modeling, which elements are identified to create an end-to-end risk picture?

• A. Actors, goals, methods, and required controls
• B. Network topology and routing policies
• C. User interface trends and branding
• D. Financial impact and budget allocations

Answer: (B)

In threat modeling for AI systems, you create an end-to-end risk picture by identifying who could threaten the system, what their objectives are, how they might achieve them, and which controls are needed to mitigate those risks. That means mapping the actors (potential adversaries or misusers), their goals (e.g., data theft, model manipulation, loss of integrity), the methods they could use (attack techniques, misuse scenarios), and the controls required (data protection, access controls, monitoring, response plans) across the entire AI lifecycle from data input to model deployment and inference.

Network topology and routing policies focus on how data flows and is protected within the network, which informs security design but doesn’t by itself describe the full, end-to-end risk landscape of an AI threat model. User interface trends and branding don’t address security threats or mitigations in the AI pipeline. Financial impact and budget allocations pertain to risk valuation and program management rather than identifying threat actors, their goals, and the concrete controls needed to reduce risk.