最新的ISACA Advanced in AI Security Management (AAISM) - AAISM免費考試真題

A. Model inversion or membership inference attacks

B. Data poisoning to corrupt training outcomes

C. Algorithmic bias leading to discriminatory lending

D. Resources needed to retrain and validate the AI model

A. Providing clear, granular, and revocable consent options to customers

B. Encrypting all data used by the platform before processing

C. Logging all interactions and storing records to service level agreements (SLAs)

D. Implementing role-based access control (RBAC) and monitoring activities

A. It ensures that scenarios are reviewed thoroughly.

B. It ensures the output is representative of target end users.

C. It reduces the perception of bias by the media and users.

D. It reduces the likelihood of a risk being overlooked.

A. Assess risk levels based on risk appetite and regulatory requirements

B. Identify vulnerabilities related to the technologies in use

C. Ensure the AI technologies are included in the asset inventory

D. Begin grouping similar AI products and solutions together

A. Maximizing the amount of training data

B. Tuning algorithms used in the AI model

C. Selecting the appropriate training data

D. Maximizing neural network size

A. Complying with applicable AI-related regulations

B. Ensuring ethical use of AI technologies in projects

C. Protecting enterprise information used by AI projects

D. Establishing an AI governance committee

A. Risk exposure due to bias in AI outputs is kept within an acceptable range

B. Management is prepared to disclose AI system architecture to stakeholders

C. Responsibility is defined for legal actions related to AI regulatory requirements

D. Ethical standards are incorporated into security awareness programs

A. Obtain data subject consent on the end user interface.

B. Conduct a review of applicable data protection regulations.

C. Have the vendor sign a nondisclosure agreement.

D. Apply encryption to safeguard personnel data.

A. Failover to a secondary environment and retrain the model with recent data.

B. Keep the system offline and conduct a full forensic investigation.

C. Require customer service to manually price and approve orders until the AI system is fully restored.

D. Revert to the previous AI model and validate backup of the feature store.

A. Establishing and monitoring acceptable levels of AI system risk

B. Creating and maintaining an AI risk register

C. Performing an AI impact assessment

D. Ensuring effective AI key performance indicators (KPIs)

A. To prevent model drift by restricting personal use of training datasets

B. To improve inference speed by controlling which APIs can be accessed

C. To ensure employees do not expose sensitive data to external AI tools

D. To reduce storage costs by limiting employee use of large models

A. Verification

B. Testing

C. Evaluation

D. Validation

A. Retrieval-augmented generation (RAG)

B. Large language model (LLM)

C. Natural language generation (NLG)

D. Natural language processing (NLP)

A. They automatically fine-tune AI models

B. They help developers create synthetic data

C. They visualize AI model performance

D. They document training data and AI model use cases