AWS Certified AI Practitioner: Exam Domains and Topics, Explained

If you want to prepare efficiently for the AWS Certified AI Practitioner (AIF-C01), start from its structure. The exam is built from five domains, and knowing what each one tests, and how much it is worth, is the fastest way to spend your study time where it counts. This is an independent breakdown based on the publicly published AWS exam guide.

Domain 1: Fundamentals of AI and ML (about 20 percent)

The foundation. You are expected to define AI, machine learning, and deep learning and tell them apart, describe the machine learning lifecycle from data to deployment, and recognize common problem types: classification, regression, clustering, and anomaly detection. Expect to map a described scenario to the right category.

Domain 2: Fundamentals of Generative AI (about 24 percent)

The heart of what makes this a modern AI exam. You need the vocabulary of generative models: tokens, embeddings, context windows, prompts and completions, temperature, and what a foundation model is. You should be able to explain, in plain language, how a large language model produces text and where its limits (such as hallucination) come from.

Domain 3: Applications of Foundation Models (about 28 percent)

The largest domain, and the one that most often decides a pass. It tests judgment: given a goal, which technique fits best?

• Prompt engineering - shaping inputs to get better outputs, including few-shot examples and clear instructions.
• Retrieval-augmented generation (RAG) - grounding a model in your own data so answers are current and specific, without retraining.
• Fine-tuning - adapting a model to a domain or style when prompting and RAG are not enough, and understanding its cost.
• Evaluation - judging model output quality, and why that is harder than it looks.

The recurring question here is not 'what is RAG' but 'in this situation, would you prompt, retrieve, or fine-tune, and why?' Practice choosing, not just defining.

Domain 4: Guidelines for Responsible AI (about 14 percent)

Smaller, but rich in easy points. You should be able to discuss fairness and bias, transparency and explainability, and the real trade-offs between them (a more accurate model can be harder to explain, for example). The exam wants to see that you treat responsible AI as an engineering concern, not a slogan.

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

How AI workloads are kept safe and accountable: protecting training and inference data, controlling access, and governing how models are used. If you have any cloud fundamentals background, much of this will feel familiar, applied to the specific case of AI systems.

Turn the map into a plan

Knowing the domains is step one. Knowing which ones you are weak in is step two, and it is the one people skip. A free diagnostic scores you across all five domains and shows you exactly where the gaps are, so your study time goes to the topics that will actually move your score.