AIF-C01 Exam Topics: What's Actually on the AWS AI Practitioner Test

AIF-C01 Exam at a Glance

If you're trying to figure out exactly what's covered on the AWS Certified AI Practitioner exam, you've come to the right place. The AIF-C01 exam topics are spread across five domains, and knowing the weight of each domain is the single most important factor in building an efficient study plan. This article breaks down every domain, explains what AWS actually tests, and tells you where to focus your limited study time.

Before diving into the domains, here are the fast facts you need to know about the AIF-C01 certification:

AWS launched the AIF-C01 in 2024 as its first dedicated AI-focused foundational certification. It requires no prerequisites - meaning anyone can sit for it, whether you're a developer, a business analyst, a project manager, or someone completely new to cloud technology. The exam costs $100 and sits at the foundational level of the AWS AI certification path from AI Practitioner to ML Specialty.

Domain 1: Fundamentals of AI and ML (20%)

Domain 1 covers the foundational concepts of artificial intelligence and machine learning that underpin everything else on the exam. At 20%, it represents roughly 13 questions out of 65. Don't underestimate it - this domain sets the vocabulary and conceptual framework you'll need to answer questions across all the other domains.

Key Topics in Domain 1

• Core ML concepts: Supervised learning, unsupervised learning, reinforcement learning, and the difference between classification, regression, and clustering tasks
• Training and inference: Understanding how models are trained on data and how inference differs from training
• Model evaluation metrics: Accuracy, precision, recall, F1 score, AUC-ROC - you need to know what each metric measures and when to use it
• Bias and variance: Overfitting, underfitting, and techniques like regularization and cross-validation
• Data concepts: Feature engineering, data preprocessing, training/validation/test splits
• AI use cases: Computer vision, natural language processing (NLP), speech recognition, recommendation systems
• AWS AI services overview: Amazon Rekognition, Amazon Comprehend, Amazon Translate, Amazon Polly, Amazon Transcribe, Amazon Forecast, and Amazon Personalize

The exam tests conceptual understanding here, not deep mathematical knowledge. You don't need to know how to implement a neural network - you need to know which type of ML problem requires which approach, and which AWS service is the right tool for a given use case.

Domain 2: Fundamentals of Generative AI (24%)

Domain 2 is the second largest section of the exam at 24%, accounting for approximately 15-16 questions. This is pure generative AI territory - large language models, foundation models, prompt engineering basics, and the mechanics of how modern AI systems generate text, images, and other outputs.

Key Topics in Domain 2

• Large Language Models (LLMs): How they work, what tokens are, context windows, temperature settings, and model parameters
• Foundation models vs traditional ML: Understanding why foundation models are different from task-specific models and why they matter
• Transformer architecture (conceptual): Attention mechanisms at a high level - you don't need to code one, but you should understand what makes transformers different
• Generative AI output types: Text, images, audio, code, embeddings - what each type is and what use cases it serves
• Prompt engineering fundamentals: Zero-shot, few-shot, and chain-of-thought prompting - what each technique achieves and when to use it
• Hallucinations and limitations: Why LLMs hallucinate, what factual grounding means, and how Retrieval-Augmented Generation (RAG) addresses these issues
• AWS Bedrock overview: What Amazon Bedrock is, which models are available through it (Anthropic Claude, Meta Llama, Stability AI, Amazon Titan, etc.), and how it's accessed

This domain is where candidates who've actually used tools like ChatGPT or Claude have an advantage - conceptual familiarity goes a long way. However, you still need to understand the AWS-specific angle, particularly around Amazon Bedrock. For a deeper look at Bedrock specifically, check out our AWS Bedrock, SageMaker and AI Services: AIF-C01 Study Notes.

Domain 3: Applications of Foundation Models (28%)

Domain 3 is the largest section of the entire exam at 28% - roughly 18 questions. This is where theoretical knowledge meets practical application. AWS wants to know that you can apply generative AI tools to real business problems using AWS services. This domain is heavily focused on Amazon Bedrock and Amazon SageMaker.

Key Topics in Domain 3

• Retrieval-Augmented Generation (RAG): How RAG works, when to use it, and how Amazon Bedrock Knowledge Bases enable RAG workflows
• Fine-tuning vs prompt engineering: When to fine-tune a model versus when a good prompt is sufficient, and the cost/complexity trade-offs involved
• Agents for Amazon Bedrock: What AI agents are, how they use tools and APIs to complete multi-step tasks, and how Bedrock Agents are configured
• Embeddings and vector databases: What embeddings are, how semantic search works, and how vector stores integrate with Bedrock
• Model evaluation: How to evaluate foundation model outputs - human evaluation, automated metrics like ROUGE and BLEU, and Amazon Bedrock Model Evaluation
• Cost optimization: How to select the right model size, use caching, and optimize inference costs in production
• SageMaker for ML workflows: SageMaker Studio, SageMaker Pipelines, SageMaker Jumpstart - at a conceptual level
• Guardrails for Amazon Bedrock: Content filtering, topic denial, PII redaction, and how guardrails enforce safe AI outputs

Because Domain 3 makes up more than a quarter of the test, your AWS AI Practitioner exam prep should allocate proportionally more time here. Working through free AIF-C01 practice questions focused on Bedrock scenarios is one of the most effective ways to prepare for this domain.

Domain 4: Guidelines for Responsible AI (14%)

Domain 4 covers responsible AI at 14% of the exam - about 9 questions. This domain has grown in importance as AI regulation has accelerated globally, and AWS has baked responsible AI principles into many of its services. The questions here are often conceptual and scenario-based rather than purely technical.

Key Topics in Domain 4

• Fairness and bias in AI: Types of bias (data bias, algorithmic bias, sampling bias), how bias enters ML pipelines, and strategies to detect and mitigate it
• Transparency and explainability: What model explainability means, why it matters, and how tools like Amazon SageMaker Clarify help
• Accountability: Human oversight of AI systems, audit trails, and shared responsibility for AI outcomes
• Privacy: Data minimization, consent, and how AWS services protect user data in AI workloads
• AWS Responsible AI principles: AWS's published framework for building responsible AI - fairness, explainability, robustness, privacy, security, and controllability
• Regulatory landscape: High-level awareness of AI regulations like the EU AI Act and what "high-risk AI" means in regulatory terms

Candidates who've worked in data governance or compliance roles often find Domain 4 intuitive. For everyone else, the key is understanding how AWS tools - particularly SageMaker Clarify, Bedrock Guardrails, and AWS Audit Manager - support responsible AI implementation.

Domain 5: Security, Compliance, and Governance for AI Solutions (14%)

Domain 5 also carries 14% of the exam weight, representing about 9 questions. If you've studied for other AWS certifications like the Cloud Practitioner or Solutions Architect, some of this will feel familiar - but it's specifically applied to AI workloads and has unique considerations.

Key Topics in Domain 5

• IAM for AI services: How Identity and Access Management (IAM) roles, policies, and permissions apply to Bedrock, SageMaker, and other AI services
• Data encryption: Encryption at rest and in transit for training data, model artifacts, and inference outputs
• Network security: VPCs, private endpoints, and PrivateLink for keeping AI workloads off the public internet
• Compliance frameworks: How AWS services map to compliance standards (SOC 2, HIPAA, GDPR) in AI contexts
• Model governance: Model versioning, lineage tracking, and audit trails using SageMaker Model Registry and AWS CloudTrail
• Prompt injection and adversarial inputs: Awareness of AI-specific security threats and how Bedrock Guardrails help mitigate them
• AWS shared responsibility model applied to AI: What AWS manages vs. what customers are responsible for when using managed AI services

What the Actual Questions Look Like

The AIF-C01 exam uses two question formats: multiple choice (one correct answer from four options) and multiple response (two or more correct answers from five options). Multiple response questions are harder because partial credit is not awarded - you must select all correct answers.

Most questions are scenario-based. AWS presents a business situation and asks you to identify the best AWS service, design approach, or course of action. Here's what a typical AIF-C01 exam question pattern looks like across each domain:

How to Study Each Domain Effectively

Knowing the AIF-C01 exam topics is only half the battle - you need a study approach that matches how AWS actually writes its questions. Here's what works:

Use the Official Exam Guide First

AWS publishes a free exam guide for AIF-C01 on its certification page. Download it. This document lists every task statement within each domain and tells you exactly what skills are being tested. Your entire study plan should be built around it. Our complete AIF-C01 study guide expands on each task statement with explanations and study tips.

Practice With Scenario-Based Questions

Because the AIF-C01 is scenario-driven, flashcards alone won't get you to 700. You need to practice reading scenarios and selecting the best answer under time pressure. Visit our practice test platform for full-length AIF-C01 practice exams that simulate the real testing experience.

Hands-On Time With Bedrock and SageMaker

Given that Domain 3 alone is 28% of the exam and centers on Bedrock and SageMaker, getting hands-on experience dramatically improves retention. AWS offers a free tier for many services. Spend an afternoon in the Bedrock playground testing different models and prompts - the conceptual knowledge will crystallize much faster.

AIF-C01 vs Other AWS Certifications

One of the most common questions candidates ask is how the AWS AI Practitioner compares to other AWS exams. The table below shows the key differences. For a much deeper comparison, read our full article on AWS AI Practitioner vs Cloud Practitioner: Which One First?

Feature	AWS AI Practitioner (AIF-C01)	AWS Cloud Practitioner (CLF-C02)	AWS ML Specialty (MLS-C01)
Level	Foundational	Foundational	Specialty
Fee	$100	$100	$300
Questions	65	65	65
Time Limit	90 minutes	90 minutes	180 minutes
Passing Score	700/1000	700/1000	750/1000
Prerequisites	None	None	Recommended 2+ years ML experience
Primary Focus	AI/ML concepts + Generative AI + Bedrock	Cloud fundamentals + AWS services broadly	Deep ML engineering + SageMaker
Best For	Anyone wanting AI credentials	Cloud beginners and business roles	ML engineers and data scientists

If you're wondering whether the certification is worth pursuing at all, our dedicated article on whether the AWS AI Practitioner certification is worth it digs into salary data and career outcomes. The short answer: for most professionals working with or adjacent to AI teams, it absolutely is.

To understand where this certification fits in your longer-term professional development, our guide on the AWS AI certification path from AI Practitioner to ML Specialty explains exactly how AIF-C01 ladders into more advanced credentials and which path makes sense for different career goals.

Ready to test your knowledge across all five domains? Head over to our free AIF-C01 practice test platform where you can attempt full-length mock exams organized by domain. It's the fastest way to identify your weak spots before exam day.

Frequently Asked Questions