AWS Certified AI Practitioner (AIF-C01) Study Guide: Exam Domains and Study Plan
A practical AWS AIF-C01 study guide covering the five exam domains, key concepts, and a focused preparation plan.
The AWS Certified AI Practitioner (AIF-C01) is for people who want to understand core AI concepts, common use cases, and AWS AI tools. The exam focuses on choosing an appropriate approach and recognizing its limits and risks. You do not need to train models from scratch or derive advanced mathematics.
This guide explains what the exam covers, which ideas are easy to confuse, and how to plan your study. Exam objectives and service lists can change, so check the latest AWS exam guide as you prepare.
Who is this exam for?
- Cloud and application developers who want to understand AI capabilities and when a tool fits a problem.
- Data and business analysts who need to know what AI can and cannot do, and how to evaluate its results.
- Technical leads and solution consultants who take part in cost, security, compliance, and risk decisions.
This is not an ML engineering or AI research certification. If you want to build and operate ML models, see MLA-C01. If you want to build production generative AI applications, see AIP-C01.
What background do you need?
You do not need coding, advanced math, or model-training experience. Basic cloud concepts help, such as accounts, permissions, and what cloud services are for. If those ideas are new, you can start with the CLF-C02 guide, but it is not a prerequisite.
You do not need to memorize every AI term in isolation. Focus on explaining what kind of problem you have, what data it needs, how you would judge the result, and what risks it may create.
Exam domains and weights
The table summarizes the five domains in the AWS guide. Weights apply to scored content. Objectives and in-scope services can change, so check the official guide again before your exam.
| Domain | Official name | Focus | Weight |
|---|---|---|---|
| 1 | Fundamentals of AI and ML | AI/ML basics, use cases, and the AI/ML lifecycle | 20% |
| 2 | Fundamentals of GenAI | Generative AI, foundation models, prompts and context, cost, and agentic AI concepts | 24% |
| 3 | Applications of Foundation Models | FM applications, customization, prompt management, evaluation, and iteration | 28% |
| 4 | Guidelines for Responsible AI | Fairness, transparency, privacy, bias, and human oversight | 14% |
| 5 | Security, Compliance, and Governance for AI Solutions | AI security, compliance, governance, output validation, and data protection | 14% |
Domain 3 has the largest weight, but do not skip the others. The fundamentals in Domains 1 and 2 support application questions. Domains 4 and 5 cover responsible and secure use.
How to study each domain
Domain 1: Fundamentals of AI and ML
Understand how AI, machine learning (ML), deep learning, and generative AI relate. Recognize common use cases such as classification, forecasting, and recommendations. Learn the AI/ML lifecycle: preparing data, training, inference, evaluation, and monitoring. You should also distinguish inference types and decide when traditional ML or a foundation model (FM, a large pretrained model that can support many tasks) is a better fit.
Common mix-up: treating every prediction as generative AI, or focusing on model scores while ignoring business results. Use an example to explain the input data, model output, and how you would measure success.
Domain 2: Fundamentals of GenAI
Learn what generative AI can and cannot do, how to select a model, and how prompts and context affect its results. The revised objectives also cover how token-based pricing affects inference cost and performance, plus agentic AI basics such as tool use, memory, multi-agent collaboration, and workflow orchestration.
Common mix-up: assuming a larger model is always more accurate, or treating retrieval-augmented generation (RAG, which grounds answers in external information) and fine-tuning (which adjusts model behavior using examples) as the same approach. Compare them by asking whether you need new knowledge or more consistent behavior.
Domain 3: Applications of Foundation Models
Study how to use models in real tasks: prompts and context, customization options such as RAG and fine-tuning, their cost trade-offs, agents for multi-step tasks, prompt version management, and ways to evaluate results with human review, benchmark data, and business metrics.
Common mix-up: choosing fine-tuning before checking whether prompts and data are enough; judging a demo without looking at failures, latency, cost, or user satisfaction. For each scenario, outline the goal, data, approach, evaluation, and next iteration.
Domain 4: Guidelines for Responsible AI
Understand fairness, transparency, explainability, privacy, and safe use. Consider where bias can come from and when a process needs human oversight, user feedback, or clear disclosure.
Common mix-up: assuming the model provider is responsible for every outcome. For each risk, name a practical control, who owns it, and when it happens.
Domain 5: Security, Compliance, and Governance
Study permissions and data protection, audit logs, threat protection, prompt injection, data leakage, output filtering, and validation. Understand how governance requirements limit the use of data and models. The AWS guide also lists relevant services; check its latest version for the current service scope.
Common mix-up: thinking only about encryption while overlooking least privilege, logging, output checks, and incident response. Review a scenario by asking who can access the data, how it is protected, what interactions are recorded, and how issues are handled.
Suggested study order
- Learn AI/ML fundamentals and common use cases (Domain 1).
- Study generative AI, foundation models, context, and agentic AI (Domain 2).
- Move to application patterns, cost trade-offs, and evaluation (Domain 3).
- Review responsible AI, security, compliance, and governance (Domains 4–5).
Build a basic framework, then check each objective in the official guide. Do not just memorize service names: know what problem a service addresses and where it fits.
A three-step preparation plan
- Use the official guide. Read the latest exam guide and mark each objective as familiar, uncertain, or new. It defines the scope, but it is not a complete question bank.
- Make concise notes. For each domain, record key ideas, common contrasts, and one practical example. Pay special attention to agents, context, cost, and evaluation in the revised objectives.
- Practice and review misses. Group mistakes by domain. Decide whether the cause was a concept gap, a poor scenario choice, or a missed risk, then revisit the relevant official material.
Check your readiness with three questions: Can you explain ML vs generative AI to a non-technical colleague? Can you compare RAG and fine-tuning for a knowledge Q&A use case? Can you name security and responsible-use checks before launch? If not, review those concepts before adding more practice questions.
Frequently asked questions
Do I need ML experience for AIF-C01?
No. Basic cloud knowledge helps, but you do not need to be an ML engineer first.
Do I have to take CLF-C02 first?
No. If cloud services and permissions are unfamiliar, build that foundation first. Otherwise, you can prepare for AIF-C01 directly.
Should I take AIF-C01, MLA-C01, or AIP-C01?
AIF-C01 focuses on AI concepts and use-case decisions. MLA-C01 focuses on ML engineering. AIP-C01 focuses on production generative AI applications. Check the latest AWS guides for each exam’s scope.
Start practicing
When you are ready, use exam code AIF-C01 to practice by domain, then use missed questions to revisit the matching objectives and official materials.