Amazon AWS AI Practitioner AIF-C01 AI Concepts Learning Models And Terminology Practice Test

 

AIF-C01 skills 1.1 | 32 original questions

This AWS Certified AI Practitioner AIF-C01 practice test focuses on ai concepts learning models and terminology through original scenario-based questions aligned to AWS Exam Guide version 1.1 published April 30, 2026. Use the full ExamSnap AIF-C01 collection for broader practice across all five current exam domains. For broader exam preparation, review the Amazon AWS Certified AI Practitioner AIF-C01 Exam Dumps page.

Instructions: Select the best answer for each question. Review the rationale after answering. Each distractor includes a brief explanation of why it is not the strongest fit for the stated scenario.

Question 1

For the analytics modernization at Woodgrove Bank, stakeholders need to place multi-layer neural-network methods in the correct hierarchy. Which concept, service, or technique most directly addresses this goal? The design must remain supportable after launch, but no additional feature is required beyond the stated need. The pilot uses 77 representative records from 7 approved data sources.

1. Deep learning is a subset of machine learning
2. Large language model (LLM)
3. Natural language processing (NLP)
4. Generative AI focuses on creating new content
5. Unsupervised learning

Correct answer: A

Why: Deep learning is a specialized branch of ML built around deep neural networks. It directly addresses the requirement in this scenario.

Option review:

A: Deep learning is a specialized branch of ML built around deep neural networks. It directly addresses the requirement in this scenario.

B: An LLM is a foundation model specialized in language tasks and often supports generation, summarization, reasoning, and conversation. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: NLP covers techniques for understanding, extracting information from, and generating human language. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: GenAI is characterized by producing new content such as text, images, audio, or code. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Unsupervised learning finds structure in unlabeled data, for example through clustering. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Deep learning is a subset of machine learning – Deep learning is a specialized branch of ML built around deep neural networks.

Question 2

Wide World Importers is comparing alternatives for its compliance-assistant prototype. The risk manager needs to discover patterns or groups when target labels are not provided. Which option is most appropriate while trying to meet a strict latency target? A short pilot window means the team prefers an approach that can be evaluated with clear success criteria. The first release supports 4 departments and is reviewed every 114 days.

1. Classification
2. Unsupervised learning
3. Supervised learning
4. Clustering
5. Regression

Correct answer: B

Why: Unsupervised learning finds structure in unlabeled data, for example through clustering. It directly addresses the requirement in this scenario.

Option review:

A: Classification predicts categorical labels. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Unsupervised learning finds structure in unlabeled data, for example through clustering. It directly addresses the requirement in this scenario.

C: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Clustering is an unsupervised technique that discovers groups based on similarity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Regression estimates a continuous quantity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Unsupervised learning – Unsupervised learning finds structure in unlabeled data, for example through clustering.

Question 3

An architecture review at VanArsdel Media has narrowed a forecasting initiative decision to one requirement: use systems that learn patterns from data instead of relying only on explicitly coded rules. What should the security architect select? The architecture board will reject a choice that addresses a different problem from the one described. The service has a 151-millisecond internal response target for the affected workflow.

1. Batch inference
2. Generative AI focuses on creating new content
3. Machine learning (ML)
4. Serverless inference
5. Computer vision

Correct answer: C

Why: ML is a subset of AI in which algorithms learn relationships from data to make predictions or decisions. It directly addresses the requirement in this scenario.

Option review:

A: Batch inference processes many records together and is appropriate when results can be produced on a schedule. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: GenAI is characterized by producing new content such as text, images, audio, or code. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: ML is a subset of AI in which algorithms learn relationships from data to make predictions or decisions. It directly addresses the requirement in this scenario.

D: Serverless inference automatically provisions serving capacity and is well suited to variable or infrequent traffic when supported. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Computer vision applies AI/ML to visual data such as images and video. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Machine learning (ML) – ML is a subset of AI in which algorithms learn relationships from data to make predictions or decisions.

Question 4

The risk manager at Datum Dynamics is preparing a recommendation for a customer-support modernization. The recommendation must distinguish an agent from a model that only returns a response to one prompt. Which choice is the best match? Budget has been approved for the project, but the team still wants to avoid unnecessary recurring consumption. The team is comparing 6 candidate designs after a 188-day proof of concept.

1. Machine learning (ML)
2. Deep learning
3. Artificial intelligence (AI)
4. Agentic AI adds goal-directed planning and action
5. Supervised learning

Correct answer: D

Why: Agentic systems can plan, call tools, maintain state, and execute multi-step workflows. It directly addresses the requirement in this scenario.

Option review:

A: ML is a subset of AI in which algorithms learn relationships from data to make predictions or decisions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Deep learning is an ML approach based on neural networks with many layers and is commonly used for complex image, speech, and language tasks. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: AI is the umbrella field that includes many approaches such as machine learning, reasoning, perception, and language processing. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Agentic systems can plan, call tools, maintain state, and execute multi-step workflows. It directly addresses the requirement in this scenario.

E: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Agentic AI adds goal-directed planning and action – Agentic systems can plan, call tools, maintain state, and execute multi-step workflows.

Question 5

Alpine Ski House has completed discovery for a agentic workflow trial. Before implementation, the security architect must decide how to learn a mapping from inputs to known labels or numeric targets. Which choice best satisfies that requirement? The team will validate the result with representative production examples before rollout. The control owner requires evidence from 3 test groups before the 225-day release review.

1. Clustering
2. Unsupervised learning
3. Reinforcement learning
4. Regression
5. Supervised learning

Correct answer: E

Why: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. It directly addresses the requirement in this scenario.

Option review:

A: Clustering is an unsupervised technique that discovers groups based on similarity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Unsupervised learning finds structure in unlabeled data, for example through clustering. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Regression estimates a continuous quantity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. It directly addresses the requirement in this scenario.

Learning point: Supervised learning – Supervised learning trains on labeled examples and is used for tasks such as classification and regression.

Question 6

While planning a contact-center transformation, Humongous Insurance identifies this requirement: identify the layered computational model composed of interconnected units with learned weights. Which option should the risk manager prioritize if the goal is to limit exposure of sensitive data? The pilot has representative data, and the team will measure the selected approach against an agreed acceptance threshold. The project has 8 downstream consumers and a monthly review of approximately 262 sampled interactions.

1. Neural network
2. Computer vision
3. Batch inference
4. Reinforcement learning
5. A foundation model can support multiple downstream tasks

Correct answer: A

Why: A neural network transforms inputs through connected layers whose parameters are learned during training. It directly addresses the requirement in this scenario.

Option review:

A: A neural network transforms inputs through connected layers whose parameters are learned during training. It directly addresses the requirement in this scenario.

B: Computer vision applies AI/ML to visual data such as images and video. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Batch inference processes many records together and is appropriate when results can be produced on a schedule. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Foundation models are trained broadly and can be adapted or prompted for many downstream tasks. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Neural network – A neural network transforms inputs through connected layers whose parameters are learned during training.

Question 7

A proof of concept at Graphic Design Institute exposed a design decision for the security architect: the solution must distinguish a data-driven learning approach from the broader AI field. Which option most directly solves that problem? The team will document the rationale for auditors and wants the recommendation to be defensible from the scenario facts. The rollout spans 5 application teams, each using the same approved requirement set for the next 299 days.

1. Generative AI focuses on creating new content
2. Machine learning is a subset of AI
3. Deep learning is a subset of machine learning
4. Generative AI (GenAI)
5. Fairness

Correct answer: B

Why: AI is the broader discipline; ML is one family of AI techniques that learns patterns from data. It directly addresses the requirement in this scenario.

Option review:

A: GenAI is characterized by producing new content such as text, images, audio, or code. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: AI is the broader discipline; ML is one family of AI techniques that learns patterns from data. It directly addresses the requirement in this scenario.

C: Deep learning is a specialized branch of ML built around deep neural networks. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: GenAI produces new content rather than only classifying or predicting a fixed label. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Fairness concerns equitable treatment and outcome patterns across users or groups. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Machine learning is a subset of AI – AI is the broader discipline; ML is one family of AI techniques that learns patterns from data.

Question 8

Relecloud is documenting the target state for a sales-assistant rollout. The risk manager needs a solution that can learn a policy by receiving rewards or penalties from interactions with an environment. Which option is the strongest fit? The team wants the least complex technically correct choice that satisfies the requirement. The evaluation set contains examples from 2 business workflows and 336 recent production cases.

1. Supervised learning
2. Classification
3. Reinforcement learning
4. Unsupervised learning
5. Clustering

Correct answer: C

Why: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. It directly addresses the requirement in this scenario.

Option review:

A: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Classification predicts categorical labels. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. It directly addresses the requirement in this scenario.

D: Unsupervised learning finds structure in unlabeled data, for example through clustering. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Clustering is an unsupervised technique that discovers groups based on similarity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Reinforcement learning – Reinforcement learning optimizes behavior from feedback signals over sequences of actions.

Question 9

Adventure Works Manufacturing is reviewing a internal search upgrade. The security architect has one primary requirement: analyze or generate human language in text or speech. Which choice best fits the requirement? The workload has passed basic feasibility checks, so the remaining question is which approach best matches the requirement. The initial rollout covers 373 internal users across 7 business units.

1. Fairness
2. Agentic AI
3. Generative AI focuses on creating new content
4. Natural language processing (NLP)
5. Batch inference

Correct answer: D

Why: NLP covers techniques for understanding, extracting information from, and generating human language. It directly addresses the requirement in this scenario.

Option review:

A: Fairness concerns equitable treatment and outcome patterns across users or groups. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Agentic AI combines models with planning, memory, tools, and orchestration to act toward goals. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: GenAI is characterized by producing new content such as text, images, audio, or code. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: NLP covers techniques for understanding, extracting information from, and generating human language. It directly addresses the requirement in this scenario.

E: Batch inference processes many records together and is appropriate when results can be produced on a schedule. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Natural language processing (NLP) – NLP covers techniques for understanding, extracting information from, and generating human language.

Question 10

During a design review for Proseware Services, the risk manager must distinguish content generation from predictive classification or clustering. The team also wants to reduce manual handling. What should the team choose? Stakeholders have ruled out a broad redesign and want the choice that most precisely addresses the stated need. The workload processes about 410 requests during its busiest hour and has a documented fallback path.

1. Agentic AI adds goal-directed planning and action
2. Machine learning (ML)
3. Reinforcement learning
4. Artificial intelligence (AI)
5. Generative AI focuses on creating new content

Correct answer: E

Why: GenAI is characterized by producing new content such as text, images, audio, or code. It directly addresses the requirement in this scenario.

Option review:

A: Agentic systems can plan, call tools, maintain state, and execute multi-step workflows. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: ML is a subset of AI in which algorithms learn relationships from data to make predictions or decisions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: AI is the umbrella field that includes many approaches such as machine learning, reasoning, perception, and language processing. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: GenAI is characterized by producing new content such as text, images, audio, or code. It directly addresses the requirement in this scenario.

Learning point: Generative AI focuses on creating new content – GenAI is characterized by producing new content such as text, images, audio, or code.

Question 11

Lucerne Publishing is moving a knowledge-assistant rollout from pilot to production. The key decision is how to discover patterns or groups when target labels are not provided. Which option is the strongest fit if the team wants to use current managed AWS capabilities? Operational ownership is already assigned, so the team is comparing technical fit rather than staffing models. The pilot uses 447 representative records from 9 approved data sources.

1. Unsupervised learning
2. Reinforcement learning
3. Supervised learning
4. Clustering
5. Classification

Correct answer: A

Why: Unsupervised learning finds structure in unlabeled data, for example through clustering. It directly addresses the requirement in this scenario.

Option review:

A: Unsupervised learning finds structure in unlabeled data, for example through clustering. It directly addresses the requirement in this scenario.

B: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Clustering is an unsupervised technique that discovers groups based on similarity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Classification predicts categorical labels. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Unsupervised learning – Unsupervised learning finds structure in unlabeled data, for example through clustering.

Question 12

A workshop at Lamna Healthcare focuses on a single decision: how to use an already trained model to produce a prediction or generated response for new input. Which option should the risk manager recommend? Existing application interfaces can accommodate any of the listed choices, so functional fit is the deciding factor. The first release supports 6 departments and is reviewed every 484 days.

1. Machine learning (ML)
2. Inference
3. Batch inference
4. Asynchronous inference
5. Bias

Correct answer: B

Why: Inference is the execution phase in which a trained model processes new data. It directly addresses the requirement in this scenario.

Option review:

A: ML is a subset of AI in which algorithms learn relationships from data to make predictions or decisions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Inference is the execution phase in which a trained model processes new data. It directly addresses the requirement in this scenario.

C: Batch inference processes many records together and is appropriate when results can be produced on a schedule. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Asynchronous inference decouples request submission from result retrieval for longer-running workloads. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Bias is systematic error or skew that can affect predictions and fairness. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Inference – Inference is the execution phase in which a trained model processes new data.

Question 13

For the personalization program at Contoso Retail, stakeholders need to distinguish broad pre-trained models from narrow task-specific models. Which concept, service, or technique most directly addresses this goal? Assume the required AWS capabilities are available in the selected Region and normal governance controls are in place. The service has a 521-millisecond internal response target for the affected workflow.

1. Large language model (LLM)
2. Unsupervised learning
3. A foundation model can support multiple downstream tasks
4. Deep learning is a subset of machine learning
5. Artificial intelligence (AI)

Correct answer: C

Why: Foundation models are trained broadly and can be adapted or prompted for many downstream tasks. It directly addresses the requirement in this scenario.

Option review:

A: An LLM is a foundation model specialized in language tasks and often supports generation, summarization, reasoning, and conversation. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Unsupervised learning finds structure in unlabeled data, for example through clustering. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Foundation models are trained broadly and can be adapted or prompted for many downstream tasks. It directly addresses the requirement in this scenario.

D: Deep learning is a specialized branch of ML built around deep neural networks. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: AI is the umbrella field that includes many approaches such as machine learning, reasoning, perception, and language processing. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: A foundation model can support multiple downstream tasks – Foundation models are trained broadly and can be adapted or prompted for many downstream tasks.

Question 14

Fourth Coffee is comparing alternatives for its developer-productivity pilot. The risk manager needs to learn a policy by receiving rewards or penalties from interactions with an environment. Which option is most appropriate while trying to meet a strict latency target? The review committee wants a direct mapping from the requirement to the chosen capability. The team is comparing 8 candidate designs after a 558-day proof of concept.

1. Classification
2. Clustering
3. Unsupervised learning
4. Reinforcement learning
5. Supervised learning

Correct answer: D

Why: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. It directly addresses the requirement in this scenario.

Option review:

A: Classification predicts categorical labels. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Clustering is an unsupervised technique that discovers groups based on similarity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Unsupervised learning finds structure in unlabeled data, for example through clustering. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. It directly addresses the requirement in this scenario.

E: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Reinforcement learning – Reinforcement learning optimizes behavior from feedback signals over sequences of actions.

Question 15

An architecture review at Margie Travel has narrowed a fraud-review pilot decision to one requirement: evaluate whether model outcomes avoid unjustified disparities across relevant groups. What should the security architect select? The solution will serve multiple internal teams, so the recommendation should be reusable without changing the core requirement. The control owner requires evidence from 5 test groups before the 595-day release review.

1. Serverless inference
2. Generative AI focuses on creating new content
3. Supervised learning
4. Asynchronous inference
5. Fairness

Correct answer: E

Why: Fairness concerns equitable treatment and outcome patterns across users or groups. It directly addresses the requirement in this scenario.

Option review:

A: Serverless inference automatically provisions serving capacity and is well suited to variable or infrequent traffic when supported. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: GenAI is characterized by producing new content such as text, images, audio, or code. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Asynchronous inference decouples request submission from result retrieval for longer-running workloads. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Fairness concerns equitable treatment and outcome patterns across users or groups. It directly addresses the requirement in this scenario.

Learning point: Fairness – Fairness concerns equitable treatment and outcome patterns across users or groups.

Question 16

The risk manager at School of Fine Art is preparing a recommendation for a analytics modernization. The recommendation must place multi-layer neural-network methods in the correct hierarchy. Which choice is the best match? The decision must follow the workload characteristics rather than a preference for the largest model or newest service. The project has 2 downstream consumers and a monthly review of approximately 632 sampled interactions.

1. Deep learning is a subset of machine learning
2. Agentic AI
3. Machine learning is a subset of AI
4. Fairness
5. Artificial intelligence (AI)

Correct answer: A

Why: Deep learning is a specialized branch of ML built around deep neural networks. It directly addresses the requirement in this scenario.

Option review:

A: Deep learning is a specialized branch of ML built around deep neural networks. It directly addresses the requirement in this scenario.

B: Agentic AI combines models with planning, memory, tools, and orchestration to act toward goals. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: AI is the broader discipline; ML is one family of AI techniques that learns patterns from data. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Fairness concerns equitable treatment and outcome patterns across users or groups. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: AI is the umbrella field that includes many approaches such as machine learning, reasoning, perception, and language processing. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Deep learning is a subset of machine learning – Deep learning is a specialized branch of ML built around deep neural networks.

Question 17

Northwind Analytics has completed discovery for a compliance-assistant prototype. Before implementation, the security architect must decide how to learn a mapping from inputs to known labels or numeric targets. Which choice best satisfies that requirement? The security baseline is already defined; the decision here concerns the specific capability described in the requirement. The rollout spans 7 application teams, each using the same approved requirement set for the next 669 days.

1. Clustering
2. Supervised learning
3. Regression
4. Reinforcement learning
5. Unsupervised learning

Correct answer: B

Why: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. It directly addresses the requirement in this scenario.

Option review:

A: Clustering is an unsupervised technique that discovers groups based on similarity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. It directly addresses the requirement in this scenario.

C: Regression estimates a continuous quantity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Unsupervised learning finds structure in unlabeled data, for example through clustering. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Supervised learning – Supervised learning trains on labeled examples and is used for tasks such as classification and regression.

Question 18

While planning a forecasting initiative, Litware Financial identifies this requirement: create new text, images, audio, video, code, or other content from learned patterns. Which option should the risk manager prioritize if the goal is to limit exposure of sensitive data? The recommendation must solve the stated requirement without introducing unrelated platform complexity. The evaluation set contains examples from 4 business workflows and 706 recent production cases.

1. Machine learning (ML)
2. Deep learning
3. Generative AI (GenAI)
4. Inference
5. Neural network

Correct answer: C

Why: GenAI produces new content rather than only classifying or predicting a fixed label. It directly addresses the requirement in this scenario.

Option review:

A: ML is a subset of AI in which algorithms learn relationships from data to make predictions or decisions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Deep learning is an ML approach based on neural networks with many layers and is commonly used for complex image, speech, and language tasks. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: GenAI produces new content rather than only classifying or predicting a fixed label. It directly addresses the requirement in this scenario.

D: Inference is the execution phase in which a trained model processes new data. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: A neural network transforms inputs through connected layers whose parameters are learned during training. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Generative AI (GenAI) – GenAI produces new content rather than only classifying or predicting a fixed label.

Question 19

A proof of concept at A. Datum Research exposed a design decision for the security architect: the solution must distinguish broad pre-trained models from narrow task-specific models. Which option most directly solves that problem? The design must remain supportable after launch, but no additional feature is required beyond the stated need. The initial rollout covers 743 internal users across 9 business units.

1. Supervised learning
2. Unsupervised learning
3. Deep learning
4. A foundation model can support multiple downstream tasks
5. Machine learning is a subset of AI

Correct answer: D

Why: Foundation models are trained broadly and can be adapted or prompted for many downstream tasks. It directly addresses the requirement in this scenario.

Option review:

A: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Unsupervised learning finds structure in unlabeled data, for example through clustering. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Deep learning is an ML approach based on neural networks with many layers and is commonly used for complex image, speech, and language tasks. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Foundation models are trained broadly and can be adapted or prompted for many downstream tasks. It directly addresses the requirement in this scenario.

E: AI is the broader discipline; ML is one family of AI techniques that learns patterns from data. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: A foundation model can support multiple downstream tasks – Foundation models are trained broadly and can be adapted or prompted for many downstream tasks.

Question 20

Coho Winery is documenting the target state for a agentic workflow trial. The risk manager needs a solution that can discover patterns or groups when target labels are not provided. Which option is the strongest fit? A short pilot window means the team prefers an approach that can be evaluated with clear success criteria. The workload processes about 780 requests during its busiest hour and has a documented fallback path.

1. Classification
2. Supervised learning
3. Regression
4. Reinforcement learning
5. Unsupervised learning

Correct answer: E

Why: Unsupervised learning finds structure in unlabeled data, for example through clustering. It directly addresses the requirement in this scenario.

Option review:

A: Classification predicts categorical labels. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Regression estimates a continuous quantity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Unsupervised learning finds structure in unlabeled data, for example through clustering. It directly addresses the requirement in this scenario.

Learning point: Unsupervised learning – Unsupervised learning finds structure in unlabeled data, for example through clustering.

Question 21

Lucerne Retail is reviewing a contact-center transformation. The security architect has one primary requirement: use the broad discipline of building systems that perform tasks associated with human intelligence. Which choice best fits the requirement? The architecture board will reject a choice that addresses a different problem from the one described. The pilot uses 817 representative records from 3 approved data sources.

1. Artificial intelligence (AI)
2. Asynchronous inference
3. Unsupervised learning
4. Agentic AI
5. Fairness

Correct answer: A

Why: AI is the umbrella field that includes many approaches such as machine learning, reasoning, perception, and language processing. It directly addresses the requirement in this scenario.

Option review:

A: AI is the umbrella field that includes many approaches such as machine learning, reasoning, perception, and language processing. It directly addresses the requirement in this scenario.

B: Asynchronous inference decouples request submission from result retrieval for longer-running workloads. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Unsupervised learning finds structure in unlabeled data, for example through clustering. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Agentic AI combines models with planning, memory, tools, and orchestration to act toward goals. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Fairness concerns equitable treatment and outcome patterns across users or groups. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Artificial intelligence (AI) – AI is the umbrella field that includes many approaches such as machine learning, reasoning, perception, and language processing.

Question 22

During a design review for Tailspin Toys, the risk manager must distinguish an agent from a model that only returns a response to one prompt. The team also wants to reduce manual handling. What should the team choose? Budget has been approved for the project, but the team still wants to avoid unnecessary recurring consumption. The first release supports 8 departments and is reviewed every 854 days.

1. Machine learning (ML)
2. Agentic AI adds goal-directed planning and action
3. Neural network
4. Computer vision
5. Reinforcement learning

Correct answer: B

Why: Agentic systems can plan, call tools, maintain state, and execute multi-step workflows. It directly addresses the requirement in this scenario.

Option review:

A: ML is a subset of AI in which algorithms learn relationships from data to make predictions or decisions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Agentic systems can plan, call tools, maintain state, and execute multi-step workflows. It directly addresses the requirement in this scenario.

C: A neural network transforms inputs through connected layers whose parameters are learned during training. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Computer vision applies AI/ML to visual data such as images and video. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Agentic AI adds goal-directed planning and action – Agentic systems can plan, call tools, maintain state, and execute multi-step workflows.

Question 23

City Power and Light is moving a sales-assistant rollout from pilot to production. The key decision is how to learn a mapping from inputs to known labels or numeric targets. Which option is the strongest fit if the team wants to use current managed AWS capabilities? The team will validate the result with representative production examples before rollout. The service has a 891-millisecond internal response target for the affected workflow.

1. Clustering
2. Reinforcement learning
3. Supervised learning
4. Unsupervised learning
5. Classification

Correct answer: C

Why: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. It directly addresses the requirement in this scenario.

Option review:

A: Clustering is an unsupervised technique that discovers groups based on similarity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. It directly addresses the requirement in this scenario.

D: Unsupervised learning finds structure in unlabeled data, for example through clustering. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Classification predicts categorical labels. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Supervised learning – Supervised learning trains on labeled examples and is used for tasks such as classification and regression.

Question 24

A workshop at Consolidated Messenger focuses on a single decision: how to adjust model parameters by learning from examples before the model is used on new inputs. Which option should the risk manager recommend? The pilot has representative data, and the team will measure the selected approach against an agreed acceptance threshold. The team is comparing 2 candidate designs after a 928-day proof of concept.

1. Artificial intelligence (AI)
2. Bias
3. Fairness
4. Training
5. Machine learning is a subset of AI

Correct answer: D

Why: Training is the process that fits model parameters using data and an optimization procedure. It directly addresses the requirement in this scenario.

Option review:

A: AI is the umbrella field that includes many approaches such as machine learning, reasoning, perception, and language processing. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Bias is systematic error or skew that can affect predictions and fairness. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Fairness concerns equitable treatment and outcome patterns across users or groups. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Training is the process that fits model parameters using data and an optimization procedure. It directly addresses the requirement in this scenario.

E: AI is the broader discipline; ML is one family of AI techniques that learns patterns from data. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Training – Training is the process that fits model parameters using data and an optimization procedure.

Question 25

For the document-intelligence project at Nod Publishers, stakeholders need to distinguish content generation from predictive classification or clustering. Which concept, service, or technique most directly addresses this goal? The team will document the rationale for auditors and wants the recommendation to be defensible from the scenario facts. The control owner requires evidence from 7 test groups before the 965-day release review.

1. Deep learning
2. A foundation model can support multiple downstream tasks
3. Agentic AI adds goal-directed planning and action
4. Fairness
5. Generative AI focuses on creating new content

Correct answer: E

Why: GenAI is characterized by producing new content such as text, images, audio, or code. It directly addresses the requirement in this scenario.

Option review:

A: Deep learning is an ML approach based on neural networks with many layers and is commonly used for complex image, speech, and language tasks. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Foundation models are trained broadly and can be adapted or prompted for many downstream tasks. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Agentic systems can plan, call tools, maintain state, and execute multi-step workflows. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Fairness concerns equitable treatment and outcome patterns across users or groups. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: GenAI is characterized by producing new content such as text, images, audio, or code. It directly addresses the requirement in this scenario.

Learning point: Generative AI focuses on creating new content – GenAI is characterized by producing new content such as text, images, audio, or code.

Question 26

Fabrikam Health is comparing alternatives for its knowledge-assistant rollout. The risk manager needs to learn a policy by receiving rewards or penalties from interactions with an environment. Which option is most appropriate while trying to meet a strict latency target? The team wants the least complex technically correct choice that satisfies the requirement. The project has 4 downstream consumers and a monthly review of approximately 42 sampled interactions.

1. Reinforcement learning
2. Clustering
3. Classification
4. Regression
5. Supervised learning

Correct answer: A

Why: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. It directly addresses the requirement in this scenario.

Option review:

A: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. It directly addresses the requirement in this scenario.

B: Clustering is an unsupervised technique that discovers groups based on similarity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Classification predicts categorical labels. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Regression estimates a continuous quantity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Reinforcement learning – Reinforcement learning optimizes behavior from feedback signals over sequences of actions.

Question 27

An architecture review at Wingtip Logistics has narrowed a claims-processing redesign decision to one requirement: use an AI system that can plan steps, choose tools, take actions, and iterate toward a goal. What should the security architect select? The workload has passed basic feasibility checks, so the remaining question is which approach best matches the requirement. The rollout spans 9 application teams, each using the same approved requirement set for the next 79 days.

1. Asynchronous inference
2. Agentic AI
3. Generative AI focuses on creating new content
4. Machine learning is a subset of AI
5. Supervised learning

Correct answer: B

Why: Agentic AI combines models with planning, memory, tools, and orchestration to act toward goals. It directly addresses the requirement in this scenario.

Option review:

A: Asynchronous inference decouples request submission from result retrieval for longer-running workloads. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Agentic AI combines models with planning, memory, tools, and orchestration to act toward goals. It directly addresses the requirement in this scenario.

C: GenAI is characterized by producing new content such as text, images, audio, or code. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: AI is the broader discipline; ML is one family of AI techniques that learns patterns from data. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Agentic AI – Agentic AI combines models with planning, memory, tools, and orchestration to act toward goals.

Question 28

The risk manager at Trey Research is preparing a recommendation for a personalization program. The recommendation must distinguish a data-driven learning approach from the broader AI field. Which choice is the best match? Stakeholders have ruled out a broad redesign and want the choice that most precisely addresses the stated need. The evaluation set contains examples from 6 business workflows and 116 recent production cases.

1. Supervised learning
2. Machine learning (ML)
3. Machine learning is a subset of AI
4. Fairness
5. Artificial intelligence (AI)

Correct answer: C

Why: AI is the broader discipline; ML is one family of AI techniques that learns patterns from data. It directly addresses the requirement in this scenario.

Option review:

A: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: ML is a subset of AI in which algorithms learn relationships from data to make predictions or decisions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: AI is the broader discipline; ML is one family of AI techniques that learns patterns from data. It directly addresses the requirement in this scenario.

D: Fairness concerns equitable treatment and outcome patterns across users or groups. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: AI is the umbrella field that includes many approaches such as machine learning, reasoning, perception, and language processing. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Machine learning is a subset of AI – AI is the broader discipline; ML is one family of AI techniques that learns patterns from data.

Question 29

Bellows College has completed discovery for a developer-productivity pilot. Before implementation, the security architect must decide how to discover patterns or groups when target labels are not provided. Which choice best satisfies that requirement? Operational ownership is already assigned, so the team is comparing technical fit rather than staffing models. The initial rollout covers 153 internal users across 3 business units.

1. Clustering
2. Supervised learning
3. Reinforcement learning
4. Unsupervised learning
5. Regression

Correct answer: D

Why: Unsupervised learning finds structure in unlabeled data, for example through clustering. It directly addresses the requirement in this scenario.

Option review:

A: Clustering is an unsupervised technique that discovers groups based on similarity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Unsupervised learning finds structure in unlabeled data, for example through clustering. It directly addresses the requirement in this scenario.

E: Regression estimates a continuous quantity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Unsupervised learning – Unsupervised learning finds structure in unlabeled data, for example through clustering.

Question 30

While planning a fraud-review pilot, Blue Yonder Airlines identifies this requirement: recognize a systematic tendency that can create inaccurate or unfair outcomes for particular patterns or groups. Which option should the risk manager prioritize if the goal is to limit exposure of sensitive data? Existing application interfaces can accommodate any of the listed choices, so functional fit is the deciding factor. The workload processes about 190 requests during its busiest hour and has a documented fallback path.

1. Natural language processing (NLP)
2. Real-time inference
3. Fairness
4. Training
5. Bias

Correct answer: E

Why: Bias is systematic error or skew that can affect predictions and fairness. It directly addresses the requirement in this scenario.

Option review:

A: NLP covers techniques for understanding, extracting information from, and generating human language. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Real-time inference is designed for low-latency synchronous responses. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: Fairness concerns equitable treatment and outcome patterns across users or groups. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Training is the process that fits model parameters using data and an optimization procedure. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Bias is systematic error or skew that can affect predictions and fairness. It directly addresses the requirement in this scenario.

Learning point: Bias – Bias is systematic error or skew that can affect predictions and fairness.

Question 31

A proof of concept at Woodgrove Bank exposed a design decision for the security architect: the solution must place multi-layer neural-network methods in the correct hierarchy. Which option most directly solves that problem? Assume the required AWS capabilities are available in the selected Region and normal governance controls are in place. The pilot uses 227 representative records from 5 approved data sources.

1. Deep learning is a subset of machine learning
2. Agentic AI adds goal-directed planning and action
3. Artificial intelligence (AI)
4. Deep learning
5. Fairness

Correct answer: A

Why: Deep learning is a specialized branch of ML built around deep neural networks. It directly addresses the requirement in this scenario.

Option review:

A: Deep learning is a specialized branch of ML built around deep neural networks. It directly addresses the requirement in this scenario.

B: Agentic systems can plan, call tools, maintain state, and execute multi-step workflows. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

C: AI is the umbrella field that includes many approaches such as machine learning, reasoning, perception, and language processing. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Deep learning is an ML approach based on neural networks with many layers and is commonly used for complex image, speech, and language tasks. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Fairness concerns equitable treatment and outcome patterns across users or groups. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Deep learning is a subset of machine learning – Deep learning is a specialized branch of ML built around deep neural networks.

Question 32

Wide World Importers is documenting the target state for a compliance-assistant prototype. The risk manager needs a solution that can learn a policy by receiving rewards or penalties from interactions with an environment. Which option is the strongest fit? The review committee wants a direct mapping from the requirement to the chosen capability. The first release supports 2 departments and is reviewed every 264 days.

1. Supervised learning
2. Reinforcement learning
3. Regression
4. Classification
5. Unsupervised learning

Correct answer: B

Why: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. It directly addresses the requirement in this scenario.

Option review:

A: Supervised learning trains on labeled examples and is used for tasks such as classification and regression. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

B: Reinforcement learning optimizes behavior from feedback signals over sequences of actions. It directly addresses the requirement in this scenario.

C: Regression estimates a continuous quantity. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

D: Classification predicts categorical labels. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

E: Unsupervised learning finds structure in unlabeled data, for example through clustering. This can be appropriate in another scenario, but it does not most directly satisfy the requirement described here.

Learning point: Reinforcement learning – Reinforcement learning optimizes behavior from feedback signals over sequences of actions.