A company is formulating an AI governance policy and wants to distinguish between the core dimensions of responsible AI identified by AWS and general system performance requirements. Which TWO of the following are characteristics of responsible AI? (Choose TWO.)

1 / 1
Select all that apply
CorrectA, D

Explanation

Question Overview

Select two characteristics of responsible AI.

Requirements to satisfy
  • 1「core dimensions of responsible AI」Ethical and trustworthiness perspectives such as fairness, explainability, privacy, and controllability
  • 2「characteristics of responsible AI」Exclude performance and infrastructure metrics
Per-option explanation
ACorrect

Privacy and security

This is correct. Privacy and security is the responsible AI characteristic that protects personal data handled by AI and prevents unauthorized access.

BIncorrect

High availability

This is incorrect. High availability relates to infrastructure reliability — the property of keeping systems accessible when needed. It is not a core dimension of responsible AI.

CIncorrect

Low latency

This is incorrect. Low latency is a performance characteristic related to response speed. It is not a core dimension of responsible AI.

DCorrect

Controllability

This is correct. Controllability is the responsible AI characteristic that allows humans to oversee AI behavior and intervene or shut it down when necessary.

EIncorrect

Maximizing throughput

This is incorrect. Maximizing throughput is a performance and efficiency goal aimed at increasing the amount of work processed per unit of time. It is not a core dimension of responsible AI.

Key Takeaway

Responsible AI includes multiple core dimensions identified by AWS.
· Fairness: Ensure no group is unjustly disadvantaged.
· Explainability: Understand and explain the rationale behind individual predictions.
· Transparency: Disclose training data, intended use, and limitations.
· Robustness: Maintain performance despite noise, variation, or adversarial attacks.
· Privacy and security: Protect personal data and prevent unauthorized access.
· Governance: Establish policies and oversight structures for AI use.
· Safety: Prevent harmful or dangerous outputs.
· Controllability: Allow humans to oversee behavior and intervene or shut down as needed.
These are the perspectives included in responsible AI.

Related Links