AAISM AI Technologies And Controls Practice Test 5 — 15 Questions

Question 1

When securing a large language model (LLM) using the OWASP Top 10 for LLMs, which control is most effective against 'Insecure Output Handling'?

Accepted Answer

Enforcing strict output validation and sanitization. Insecure Output Handling occurs when a model output is interpreted by a downstream system without sanitization. Output validation (canonicalization) is the primary defense.

Answer

Using fine-grained RBAC

Answer

Implementing Rate Limiting

Answer

Encrypting training data

Question 2

An AI Security analyst is configuring Google Cloud Vertex AI Model Garden to ensure compliance with organization policies. Which IAM permission is strictly required to prevent unauthorized users from deploying a base model to a public endpoint?

Accepted Answer

aiplatform.endpoints.create. The roles/aiplatform.user role allows usage, but 'aiplatform.endpoints.create' or specific restrictions on the Endpoint resource are needed to prevent public exposure.

Answer

aiplatform.metadata.read

Answer

aiplatform.jobs.create

Answer

aiplatform.models.upload

Question 3

You are configuring private connectivity for an AI model using AWS PrivateLink. Which component is required to ensure that model inference traffic never traverses the public internet?

Accepted Answer

VPC Interface Endpoint. An interface VPC endpoint (AWS PrivateLink) allows private connectivity to SageMaker endpoints from within a VPC.

Answer

Internet Gateway (IGW)

Answer

Security Group ingress rule

Answer

NAT Gateway

Question 4

In the context of poisoning attacks, which technique is most effective for detecting data contamination in an automated ML training pipeline?

Accepted Answer

Statistical outlier detection on training feature sets. Outlier detection (or anomaly detection) is the primary mechanism to identify malicious, poisoned data points that deviate from the expected distribution of the training set.

Answer

Increasing training epochs

Answer

Differential Privacy

Answer

Model quantization

Question 5

An organization uses IBM watsonx.ai to manage model governance. Which tool is primarily used to track 'FactSheets' for model lineage and auditability?

Accepted Answer

FactSheets. The IBM watsonx.ai Governance module uses FactSheets to automatically capture and document the lifecycle, lineage, and validation status of models.

Answer

AutoAI Experiment Tracker

Answer

Pipeline Orchestrator

Answer

Model Risk Management (MRM) dashboard

Question 6

You are securing a PyTorch model and want to implement differential privacy to protect the training data. Which library is specifically designed to work with PyTorch for this purpose?

Accepted Answer

Opacus. Opacus is the official PyTorch library for training models with differential privacy.

Answer

TensorFlow Privacy

Answer

Hugging Face Evaluate

Answer

Adversarial Robustness Toolbox (ART)

Question 7

What is the purpose of 'Model Signing' in an MLOps pipeline?

Accepted Answer

To ensure model integrity and non-repudiation. Model signing uses digital signatures to verify that the model artifact has not been tampered with between the training environment and the production environment.

Answer

To encrypt the model weights at rest

Answer

To restrict access to the model registry

Answer

To increase inference speed

Question 8

To monitor for AI security incidents, you are integrating logs from an AI gateway into a SIEM. Which specific field in the request log is most critical for detecting 'Model Denial of Wallet' attacks?

Accepted Answer

usage.total_tokens. Denial of Wallet attacks involve high-cost, high-token-count requests. Monitoring the 'usage.total_tokens' or 'billing_tokens' field is critical for identifying this behavior.

Answer

latency_ms

Answer

request_ip_address

Answer

model_id

Question 9

You are auditing an AI system for compliance with the EU AI Act. Which control is mandatory for 'High-Risk' AI systems to ensure transparency for human oversight?

Accepted Answer

Human-in-the-loop design and logging. The EU AI Act requires that high-risk systems be designed with human-in-the-loop (HITL) capabilities and adequate logging to ensure transparency and accountability.

Answer

Automated decision-making without oversight

Answer

Mandatory anonymization of all input data

Answer

Exclusive use of proprietary datasets

Question 10

You are using the Adversarial Robustness Toolbox (ART) to assess a classifier. Which attack technique would you simulate to test the model's vulnerability to 'Evasion' attacks?

Accepted Answer

Fast Gradient Sign Method (FGSM). The Fast Gradient Sign Method (FGSM) is a classic white-box attack technique used to test evasion vulnerability by introducing imperceptible noise.

Answer

Membership Inference

Answer

Model Inversion

Answer

Data Poisoning via label flipping

Question 11

Which TWO actions should be taken when implementing an AI security monitoring strategy in a production environment?

Accepted Answer

Implement automated alerting for anomalous request patterns. Effective monitoring requires both tracking system health and analyzing input/output traffic for malicious patterns.

Answer

Disable all API telemetry to reduce attack surface

Answer

Use a static list of IP addresses for all model access

Answer

Perform daily model retraining regardless of drift

Question 12

Which THREE security controls are essential for protecting the integrity of the AI model training supply chain?

Accepted Answer

Data provenance and lineage tracking. Supply chain security involves securing the training data, the model architecture, and the software dependencies.

Answer

Hardening the physical security of the server room

Answer

Increasing the number of neurons in the model

Question 13

Which TWO steps are recommended to secure access to an AI model registry?

Accepted Answer

Implement Role-Based Access Control (RBAC). Access control for registries should follow the principle of least privilege and use identity federation.

Answer

Store registry credentials in plain text configuration files

Answer

Disable audit logging to reduce storage costs

Answer

Allow 'Public' access for all users to promote model sharing

Question 14

Which THREE strategies are effective for defending against model poisoning in an ML pipeline?

Accepted Answer

Implementing model validation against a clean, golden dataset. Defending against poisoning requires cleaning the training data and using robust learning algorithms.

Answer

Increasing the model's learning rate to 'burn off' poison

Answer

Regularly changing the model's activation functions

Question 15

Which TWO techniques are effective for detecting and mitigating 'Model Inversion' attacks where an adversary attempts to reconstruct training data from model outputs?

Accepted Answer

Rate limiting and monitoring query frequency. Model inversion is mitigated by reducing the information content of model outputs and limiting query volume.

Answer

Disabling all model API documentation

Answer

Increasing the precision of model output scores

Answer

Using larger batch sizes during inference