AIF-C01 Practice Test 43 — 15 Questions

Question 1

A company is developing an AI system that generates news articles. To comply with transparency regulations, they must clearly indicate when content is AI-generated. Which action should they take?

Accepted Answer

Include a disclaimer in the article metadata and a visible label stating 'This content was generated by AI'. Transparency regulations (such as the EU AI Act's Article 50 and similar disclosure requirements) mandate that AI-generated content be clearly identifiable to the audience. Option D satisfies this by combining a machine-readable disclaimer in the article metadata with a human-readable visible label, covering both automated detection and end-user awareness. This dual approach is the standard compliance pattern for generative AI disclosure.

Answer

Register the model with a government agency

Answer

Remove any identifiable information about the model from the output

Answer

Use a watermark that is invisible to users

Question 2

Which of the following is a supervised learning algorithm used for classification tasks?

Accepted Answer

Logistic regression. Logistic regression is a supervised learning algorithm because it learns from labeled training data to map input features to discrete class labels. Despite its name, it is used for binary classification tasks (e.g., spam vs. not spam) by applying a logistic (sigmoid) function to output a probability between 0 and 1, then thresholding to assign a class.

Answer

Principal Component Analysis (PCA)

Answer

Autoencoders

Answer

K-means

Question 3

Which AWS service can extract text from scanned PDF documents, including handwriting and checkboxes?

Accepted Answer

Amazon Textract. Amazon Textract is specifically designed to extract text, handwriting, and form data (including checkboxes) from scanned documents and PDFs. It uses machine learning to recognize printed and handwritten text, as well as the selection state of checkboxes, making it the correct service for this use case.

Answer

Amazon Transcribe

Answer

Amazon Comprehend

Answer

Amazon Rekognition

Question 4

A company is deploying a customer‑facing chatbot using Amazon Bedrock. They need to ensure the chatbot never reveals personally identifiable information (PII) and refuses to discuss the topic of 'employee salaries'. Which TWO Bedrock Guardrails features should they configure together? (Select TWO.)

Accepted Answer

Topic denial. Option A (Topic denial) is correct because Amazon Bedrock Guardrails' denied topics feature lets you define a custom topic such as 'employee salaries' with a name, definition, and example phrases, and the guardrail will block user inputs and model responses that fall into that topic, directly satisfying the requirement to refuse discussing salaries. Option C (PII detection and redaction) is correct because Guardrails' sensitive information filters can detect and either block or mask PII entities (such as NAME, EMAIL, PHONE, SSN, and ADDRESS) in both prompts and completions, ensuring the chatbot never reveals personally identifiable information. Option B (Grounding check) is not correct here because grounding and relevance checks only validate whether responses are supported by and relevant to a provided source (RAG) context, which does not prevent PII leakage or off-topic salary discussions. Option D (Bedrock Knowledge Base) is not correct because a knowledge base is a managed RAG resource for retrieving source content, not a guardrail feature for blocking topics or redacting PII. Option E (Content filtering) is not correct because content filters target harmful categories such as hate, violence, sexual, insults, and misconduct, not the specific business rules of PII redaction and salary-topic denial.

Answer

Grounding check

Answer

Bedrock Knowledge Base

Answer

Content filtering

Question 5

A healthcare company is building a medical diagnosis assistant using Amazon Bedrock. They need to ensure the model’s responses are based on the latest medical research and do not include outdated information. The company also wants to minimize costs. Which TWO actions should they take? (Select TWO)

Accepted Answer

Use a smaller model like Amazon Titan Text Lite to reduce inference costs. Option B is correct because Amazon Titan Text Lite is a smaller, lower-cost model, so using it for inference directly reduces per-token costs, which aligns with the company's goal to minimize costs. Option C is correct because Retrieval Augmented Generation (RAG) retrieves relevant passages from an up-to-date vector store of the latest medical journals and injects them into the prompt, grounding responses in current research and avoiding outdated information without retraining the model. Option A is not correct because the Converse API only manages conversation history and does not by itself ensure access to the latest medical research or reduce costs. Option D is not correct because monthly fine-tuning of a large model is expensive, slow, and risks stale knowledge between training cycles, conflicting with the cost-minimization goal. Option E is not correct because stuffing all research into a large context window increases token usage and cost and does not guarantee retrieval of the most current, relevant information.

Answer

Use the Converse API to maintain conversation history

Answer

Fine-tune a large model on the latest medical data monthly

Answer

Choose a model with the largest context window to include all research in the prompt

Question 6

A team is building a real-time anomaly detection system for IoT sensor data. The data is unlabeled, and the team expects the anomalies to be rare but of high importance. Which combination of approach and AWS service should the team use?

Accepted Answer

Use a semi-supervised one-class classifier with Amazon Lookout for Equipment. Amazon Lookout for Equipment is purpose-built for anomaly detection on unlabeled industrial sensor data. It uses a semi-supervised one-class classifier that learns the normal operating envelope from historical data and flags rare, high-importance anomalies in real time, matching the exact requirements of the scenario.

Answer

Use Amazon Rekognition Custom Labels to detect anomalies in sensor images

Answer

Train a supervised classifier using Amazon SageMaker; use Amazon Kinesis Data Analytics for real-time inference

Answer

Apply k-means clustering with Amazon SageMaker; deploy the model as a real-time endpoint

Question 7

A team trained a deep learning model that achieves 99% accuracy on training data but only 70% on validation data. What is the most likely issue?

Accepted Answer

Overfitting. The model performs exceptionally well on training data (99% accuracy) but significantly worse on validation data (70% accuracy). This large gap indicates the model has memorized the training data, including noise and irrelevant patterns, rather than learning generalizable features — a classic symptom of overfitting.

Answer

Underfitting

Answer

Data leakage

Answer

Feature scaling

Question 8

A data science team is evaluating foundation models for a code generation task. They need a model that is fine-tuned for code and can be deployed on Amazon SageMaker. Which THREE criteria are important to consider when selecting a model?

Accepted Answer

Licensing and usage terms. Option A (Licensing and usage terms) is correct because foundation models on SageMaker come with varying licenses (e.g., Apache 2.0, Llama Community License, or proprietary commercial terms), and the team must verify the license permits their intended commercial code-generation use case before deployment. Option B (Cost per token for inference) is correct because SageMaker endpoints bill based on instance hours and/or token throughput, so understanding per-token inference cost is essential for budgeting a production code-generation workload at scale. Option C (Context window length) is correct because code generation often requires ingesting large files or repositories, and a model's maximum context window (e.g., 4K, 8K, 32K, or 128K tokens) directly determines how much source code can be passed in a single prompt. Option D is not a primary selection criterion because the training algorithm (e.g., RLHF, SFT) is an internal implementation detail that does not directly affect deployment fit or task suitability. Option E is not a primary criterion because model size and architecture are secondary considerations already reflected in measurable factors like cost, latency, and context window, rather than being a distinct decision driver for this scenario.

Answer

The training algorithm used

Answer

Model size and architecture

Question 9

A data scientist is using Amazon SageMaker Clarify to generate a bias report for a binary classification model. They want to understand which features most influence the model's predictions. Which feature of Clarify should they use?

Accepted Answer

Feature importance from SHAP values. SageMaker Clarify provides SHAP-based feature importance values to explain model predictions. It can compute both local and global feature importance.

Answer

Partial dependence plots

Answer

Bias metrics report

Answer

Data quality report

Question 10

A machine learning practitioner is training a model to forecast product demand and observes that the model performs well on training data but poorly on unseen data. Which of the following is the MOST likely cause?

Accepted Answer

Overfitting due to excessive model complexity. The model performs well on training data but poorly on unseen data, which is the classic symptom of overfitting. Overfitting occurs when the model is excessively complex (e.g., too many parameters, deep neural networks with high capacity) and learns noise and random fluctuations in the training data rather than the underlying pattern. This leads to high variance, causing excellent training performance but poor generalization to new data.

Answer

High bias-variance tradeoff favoring bias

Answer

Underfitting due to model being too simple

Answer

Data leakage causing artificially high training scores

Question 11

A data scientist wants to deploy a custom model built with TensorFlow to Amazon SageMaker for real-time inference. Which TWO steps are required? (Choose two.)

Accepted Answer

Upload the model artifacts to an S3 bucket. Option B is correct because SageMaker requires model artifacts (the trained TensorFlow SavedModel or model.tar.gz) to be stored in an Amazon S3 bucket, which is then referenced by the SageMaker model when deploying. Option D is correct because deploying to a real-time endpoint requires creating an endpoint configuration that specifies the production variant, instance type, and initial instance count, which is then used to create the endpoint. Option A is not required because the TensorFlow inference container is already provided and maintained by SageMaker as a prebuilt Docker image, so no custom ECR repository is needed unless using a custom container. Option C is not required because the model is already built with TensorFlow; a SageMaker training job is only needed if training within SageMaker. Option E is not required because SageMaker's TensorFlow container supports native TensorFlow SavedModel format, so ONNX conversion is unnecessary.

Answer

Create an Amazon ECR repository for the inference container

Answer

Submit a training job to SageMaker

Answer

Convert the model to ONNX format

Question 12

Which TWO of the following are examples of supervised learning tasks that can be performed using Amazon SageMaker built-in algorithms?

Accepted Answer

XGBoost. XGBoost (B) is correct because Amazon SageMaker's built-in XGBoost algorithm is a supervised gradient-boosted trees implementation used for classification and regression on labeled data. Linear Learner (C) is correct because SageMaker's built-in Linear Learner algorithm trains supervised linear models for classification and regression using labeled datasets. PCA (A) is not correct because it is an unsupervised dimensionality-reduction technique that does not use target labels. LDA (D) is not correct because it is an unsupervised topic-modeling algorithm. K-Means (E) is not correct because it is an unsupervised clustering algorithm that groups unlabeled data.

Answer

Principal Component Analysis (PCA)

Answer

Latent Dirichlet Allocation (LDA)

Answer

K-Means

Question 13

A data scientist is fine-tuning a foundation model on SageMaker. They want to prevent overfitting. Which THREE actions can help? (Select THREE.)

Accepted Answer

Apply dropout. Option A (Apply dropout) is correct because randomly deactivating neurons during training acts as a regularizer that reduces the model's reliance on specific weights, thereby curbing overfitting. Option B (Increase training data size) is correct because more diverse examples give the model a broader signal and reduce the chance it memorizes the limited training set. Option D (Use early stopping) is correct because monitoring validation loss and halting training when it stops improving prevents the model from continuing to fit noise in the training data. Option C (Increase the number of epochs) is not correct because training longer typically worsens overfitting rather than preventing it. Option E (Use a smaller learning rate) is not correct because a smaller learning rate mainly affects optimization stability and convergence speed, not the model's tendency to overfit.

Answer

Increase the number of epochs

Answer

Use a smaller learning rate

Question 14

Which AWS services can be used to build, train, and deploy custom machine learning models? (Choose two.)

Accepted Answer

AWS Deep Learning AMIs. AWS Deep Learning AMIs (C) are correct because they provide pre-configured Amazon EC2 images with popular deep learning frameworks (TensorFlow, PyTorch, MXNet) and GPU drivers, giving you the full environment needed to build and train custom models on your own infrastructure. Amazon SageMaker (E) is correct because it is a fully managed platform whose built-in capabilities (notebooks, training jobs, automatic model tuning, and hosting endpoints) let you build, train, and deploy custom ML models end to end. The other options are managed AI services with pre-trained models rather than tools for creating custom models: Amazon Polly (A) only converts text to lifelike speech, Amazon Lex (B) only builds conversational chatbots using NLU, and Amazon Rekognition (D) only provides pre-trained image and video analysis.

Answer

Amazon Polly

Answer

Amazon Lex

Answer

Amazon Rekognition

Question 15

A company is training a deep learning model for image classification. Which THREE practices help reduce overfitting? (Choose three.)

Accepted Answer

L2 regularization. L2 regularization (option A) is correct because it adds a penalty proportional to the squared magnitude of the weights to the loss function, discouraging large weights and thus reducing the model's ability to memorize training noise. Dropout (option D) is correct because it randomly deactivates a fraction of neurons during each training step, forcing the network to learn redundant, more robust feature representations instead of relying on specific units. Data augmentation (option E) is correct because it synthetically expands the training set by applying label-preserving transformations (e.g., flips, rotations, crops) to images, exposing the model to more varied inputs and improving generalization. Increasing model depth (option B) is not correct here because adding layers increases capacity, which typically makes overfitting worse rather than better. Increasing the learning rate (option C) is not correct because it only affects optimization speed and stability, and an excessively high rate can cause divergence or poor convergence, not reduced overfitting.

Answer

Increasing model depth

Answer

Increasing learning rate