Question 39
Domain 5: Deployment, Scaling, Safety, and ComplianceDuring the development of an agent designed to interact with humans in a real-world environment, which development practice is most critical to ensure the agent behaves safely and predictably?
Correct answer: B
Explanation
Implementing guardrails and fail-safes is essential because an agent in a real-world environment must be constrained by the specific risks and limits of that setting. Safety-focused design reduces harmful or unpredictable actions by enforcing environment-specific constraints and fallback behavior when conditions change.
Why each option is right or wrong
A. Enabling unrestricted exploration of all possible actions
B. Implementing guardrails and fail-safes based on environment-specific constraints
The critical issue in a real-world deployment is that the agent’s action space must be bounded by the hazards and operating limits of the specific environment, so safety logic has to be built into the control layer before release. In practice, that means hard constraints, exception handling, and fallback states are used to prevent unsafe outputs or actions when sensor data is incomplete, conditions change, or the agent encounters an out-of-distribution situation.
C. Training the agent solely in simulated environments without real-world constraints
D. Disabling logging to improve agent response latency