AWS Certified AI Practitioner
Exam General Information
Official Scope and Verification
This lesson is mapped to the verified AWS Certified AI Practitioner (AIF-C01) outline. Official sources and public status were rechecked on 2026-07-13. AWS pages remain authoritative for late-breaking blueprint, availability, scheduling, price, language, delivery, and retake changes.
Official Objective Map
| Domain or objective area | Published weight | Key objective groups | Official source |
|---|---|---|---|
| Fundamentals of AI and ML | 20% | Basic AI terminology, practical AI use cases, and the AI/ML development lifecycle | AWS official AIF-C01 exam guide |
| Fundamentals of generative AI | 24% | GenAI concepts, capabilities and limitations, and AWS technologies for GenAI applications | AWS official AIF-C01 exam guide |
| Applications of foundation models | 28% | Application design, prompting, training and fine-tuning, and foundation-model evaluation | AWS official AIF-C01 exam guide |
| Guidelines for responsible AI | 14% | Responsible AI development, transparency, and explainability | AWS official AIF-C01 exam guide |
| Security, compliance, and governance for AI solutions | 14% | Securing AI systems and recognizing governance and compliance requirements | AWS official AIF-C01 exam guide |
| In-scope AWS services and features | Published without a scored percentage | Analytics, cost management, compute, containers, databases, developer tools, ML, management, networking, security, and storage | AWS official AIF-C01 in-scope services list |
Authoritative Sources for This Scope
- AWS official AIF-C01 exam guide - Official source; accessed 2026-07-13.
- AWS official AIF-C01 in-scope services list - Official source; accessed 2026-07-13.
Start here if you are learning on your own. This lesson explains what the AWS Certified AI Practitioner exam is trying to measure, what you need before you begin, how much the exam costs, what happens if you fail, and how to turn the official exam guide into a practical study plan.
All administrative facts in this lesson were checked against AWS official pages on July 13, 2026. Exam prices, policies, languages, and scheduling rules can change, so treat the AWS pages linked below as the final source before you pay for an exam.
What This Certification Proves
The AWS Certified AI Practitioner certification is a foundational credential. It is not a deep machine learning engineering exam. It validates that you can understand AI, machine learning, and generative AI concepts, recognize AWS AI services, and choose responsible, practical AI approaches for business scenarios.
Think of the exam as a service-selection and judgment test. A typical question does not ask you to code a model. It asks you to identify which AI pattern or AWS service best fits a business need, a risk constraint, a data source, or an operational requirement.
- Good fit: business analysts, product managers, project managers, sales professionals, marketing professionals, IT support staff, managers, and early technical learners who need AI/ML vocabulary and AWS service awareness.
- Also useful for: developers and data learners who want a gentle AWS AI starting point before AWS Certified Machine Learning Engineer - Associate or a deeper generative AI role.
- Not the main target: candidates who want a coding-heavy model-building exam, MLOps implementation exam, or advanced statistics exam.
What You Need To Get Started
You can begin without being an AWS engineer. The official target candidate has up to 6 months of exposure to AI/ML technologies on AWS and uses, but does not necessarily build, AI/ML solutions.
- Basic AWS cloud vocabulary. Know what services such as Amazon EC2, Amazon S3, AWS Lambda, Amazon Bedrock, and Amazon SageMaker AI are used for at a high level.
- Security basics. Understand the shared responsibility model, IAM, least privilege, encryption, and why prompts and retrieved documents can contain sensitive data.
- Pricing awareness. You do not need to memorize every price. You should understand that model choice, usage volume, tokens, storage, requests, and managed service features can affect cost.
- AI vocabulary. Be comfortable with terms such as AI, ML, deep learning, supervised learning, unsupervised learning, foundation model, LLM, embedding, token, prompt, inference, fine-tuning, RAG, hallucination, bias, and guardrail.
- An AWS Training and Certification account. Use it to schedule the exam, access your exam history, download score reports, and manage certification benefits.
- Optional AWS account for practice. Hands-on exploration helps, but keep budgets and alerts enabled. For this exam, reading service pages and doing small demos is usually enough.
If you are new to both IT and AWS Cloud, AWS recommends starting with foundational cloud training such as AWS Cloud Practitioner Essentials or AWS Technical Essentials before jumping into the AI Practitioner prep path.
Official Exam Facts
| Item | What to know | Why it matters |
|---|---|---|
| Exam code | AIF-C01 | Use the code when checking the current exam guide, scheduling page, and score report. |
| Level | Foundational | Expect breadth and concept judgment, not deep model implementation. |
| Duration | 90 minutes | For 65 questions, this is about 83 seconds per question if you spend time evenly. |
| Format | 65 questions | The exam guide says 50 questions affect your score and 15 are unscored, but unscored items are not identified. |
| Question types | Multiple choice, multiple response, ordering, and matching | Practice single-answer MCQs, but also prepare for "select all that apply," sequence, and match-the-pair reasoning. |
| Passing score | 700 on a 100-1,000 scaled score range | AWS uses scaled scoring. You do not need to pass every domain separately; the model is compensatory. |
| Cost | 100 USD | Taxes, foreign exchange, vouchers, and local payment rules can change the amount you pay at checkout. |
| Testing options | Pearson VUE test center or online proctored exam | Choose online only if your room, ID, webcam, microphone, internet, and system test are ready. |
| Validity | 3 years | Plan to recertify before expiration if you want the credential to remain active. |
Verify these facts on the AWS Certified AI Practitioner exam page and the official AIF-C01 exam guide.
Exam Cost, Retake Rules, And Scheduling Rules
Budget for the exam before you schedule. As of the AWS page checked on July 13, 2026, the listed exam cost is 100 USD. Your checkout may also reflect taxes, exchange rates, or voucher rules.
| Situation | AWS policy to remember | Self-learner action |
|---|---|---|
| You fail the exam | You must wait 14 calendar days before retaking. There is no limit on attempts, but each attempt requires the full registration fee. | Do not immediately reschedule without reviewing the score report and rebuilding weak domains. |
| You pass the exam | You cannot retake the same exam for two years, unless AWS releases a new exam guide and exam series code that makes you eligible for the new version. | Move on to the next credential or recertification path instead of trying to raise the same score. |
| You need to reschedule | AWS says you can reschedule up to 24 hours before the appointment, and each exam appointment can only be rescheduled twice. | Use reschedules sparingly. If you need a third change, AWS says you must cancel and schedule again. |
| You need to cancel | AWS states that cancellation more than 24 hours before the appointment can refund the exam fee paid at purchase. | Do not wait until the final day if your schedule is uncertain. |
| You are within 24 hours of the exam | AWS says cancellation or rescheduling is not available within 24 hours, and refunds are not available except for qualifying emergencies handled by the test delivery provider. | Run the online proctoring system test and ID check before the 24-hour window closes. |
Read the current policy pages before scheduling: AWS before-testing policies, AWS after-testing and retake policies, and AWS scheduling and testing options.
What Is In Scope
The AIF-C01 exam guide divides scored content into five domains. Use the weights to decide how much study time to spend, not just how many notes to write.
| Domain | Weight | How to study it |
|---|---|---|
| Fundamentals of AI and ML | 20% | Learn the vocabulary and use-case patterns: prediction, classification, computer vision, NLP, recommendations, training, inference, evaluation, and bias. |
| Fundamentals of GenAI | 24% | Understand foundation models, LLMs, prompts, tokens, embeddings, context windows, temperature, hallucination, prompt engineering, RAG, and fine-tuning. |
| Applications of Foundation Models | 28% | This is the largest domain. Practice matching business needs to chat, summarization, content generation, search, RAG, agents, knowledge assistants, and evaluation patterns. |
| Guidelines for Responsible AI | 14% | Study fairness, transparency, privacy, safety, explainability, human oversight, toxicity, hallucinations, and responsible use. |
| Security, Compliance, and Governance for AI Solutions | 14% | Connect IAM, encryption, logging, monitoring, data governance, compliance, and shared responsibility to AI workloads. |
The service list is not only "AI services." AWS includes analytics, compute, storage, security, governance, cost management, and database services because exam scenarios often combine AI with normal cloud controls. Use the AIF-C01 in-scope AWS services list when you want to check whether a service is fair game.
What Is Not The Main Focus
The official guide says the target candidate is not expected to perform advanced implementation tasks such as developing AI/ML models or algorithms, implementing feature engineering, tuning hyperparameters, building AI/ML pipelines, doing mathematical model analysis, or designing full governance frameworks.
This matters because self learners often over-study the wrong material. If you spend a week learning gradient descent math but cannot explain when to use Amazon Bedrock Knowledge Bases instead of fine-tuning, your study time is misallocated for AIF-C01.
How To Read AWS Exam Questions
Read each question as a set of constraints. AWS distractors often sound technically valid, but fail the scenario because they ignore scope, cost, speed, governance, data freshness, operational effort, or the user's role.
- Find the task verb. Words like identify, describe, choose, reduce, secure, evaluate, and monitor tell you what kind of answer is needed.
- Find the constraint. Look for phrases such as "least operational effort," "sensitive data," "current internal documents," "business user," "real-time," "cost-effective," or "responsible AI."
- Classify the pattern. Is this a prebuilt AI service, a foundation model app, RAG, an agent, a governance control, a monitoring concern, or a security requirement?
- Reject answers that are too heavy. Custom training, fine-tuning, or full MLOps pipelines are rarely the first answer for a foundational business scenario.
- Do not leave blanks. The exam guide says unanswered questions are scored as incorrect and there is no penalty for guessing.
Example 1: Service Selection
Scenario: A retail team wants to identify whether customer reviews are positive, negative, or neutral. They do not want to train a custom model.
Reasoning: The task is sentiment analysis. The constraint is low custom ML effort. A managed NLP service is a better match than building a model from scratch.
Exam-ready answer: Choose a managed AI service such as Amazon Comprehend for text analysis, not SageMaker training unless the scenario requires a custom model.
Example 2: RAG Versus Fine-Tuning
Scenario: An HR team wants a chatbot to answer questions from current policy documents stored in the company knowledge base. The answers must reflect policy updates quickly and cite source material.
Reasoning: The important words are "current policy documents" and "cite source material." This points to retrieval-augmented generation, not retraining the model whenever a policy changes.
Exam-ready answer: Use a retrieval-grounded pattern such as Amazon Bedrock Knowledge Bases or a managed enterprise assistant pattern, depending on the exact scenario.
Example 3: Responsible AI And Guardrails
Scenario: A public-facing assistant must avoid unsafe content, block disallowed topics, and reduce sensitive information exposure.
Reasoning: The priority is safety and control. A larger model alone does not solve policy enforcement. The answer should include application controls, testing, monitoring, and guardrails.
Exam-ready answer: Use safety controls such as Amazon Bedrock Guardrails, and pair them with clear policies, logging, human review when needed, and ongoing evaluation.
Example 4: Multiple Response Discipline
Scenario: A question asks for two controls that help protect sensitive documents used by an AI application.
Reasoning: In multiple response questions, one correct answer is not enough. You must select all correct responses. Security questions often combine identity, encryption, monitoring, and data governance.
Exam-ready answer: Likely controls could include IAM least privilege, AWS KMS encryption, CloudTrail logging, CloudWatch monitoring, Macie for sensitive data discovery, or Secrets Manager, depending on the wording. Do not choose prompt wording as a replacement for real access control.
A Practical Self-Study Plan
Use this sequence if you are preparing independently and do not know where to start.
- Day 1: orient. Read this lesson, the AWS exam page, and the official exam guide. Write down the five domains and weights from memory.
- Days 2-4: vocabulary. Study AI, ML, deep learning, GenAI, foundation models, embeddings, tokens, prompts, RAG, fine-tuning, agents, hallucination, bias, and evaluation.
- Days 5-8: service map. Build a table of in-scope services. For each service, write "input, output, best use case, common distractor."
- Days 9-11: responsible AI and security. Practice questions involving privacy, safety, guardrails, explainability, human oversight, IAM, KMS, CloudTrail, CloudWatch, and shared responsibility.
- Days 12-14: mixed practice. Use flashcards and quizzes. For every missed item, write the requirement word that made the correct option better.
- Final 48 hours: simulate pressure. Do timed sets. Practice flagging hard questions, guessing when needed, and returning after easier items.
If you have more time, stretch the same plan over 4 weeks and add small AWS console demos. If you have less time, prioritize Domain 3, Domain 2, and service-selection drills because they represent the largest share of scored content.
How To Use This Reviewer
Use the course in three passes:
- Read the lesson. Get the mental model first. Do not memorize every service feature before understanding the scenario pattern.
- Open the matching syllabus branch. Review the official domain and topic structure, then drill the related flashcards.
- Take mixed quizzes. Mix adjacent domains so you can tell similar choices apart. For example, compare Amazon Bedrock, Amazon Q, SageMaker AI, Amazon Comprehend, and Amazon Kendra in one session.
The fastest improvement comes from explaining wrong answers. A wrong option is usually attractive because it is an AWS service, but not the best service for the scenario.
Exam-Day Strategy
- Use the first pass for easy points. Answer questions you can solve confidently, flag uncertain ones, and keep moving.
- Budget time. With 90 minutes for 65 questions, aim to finish the first pass with at least 15 minutes left for flagged items.
- Watch "select all" wording. Multiple response questions require every correct option. A partly correct selection earns no credit for that item.
- Prefer managed services when the scenario asks for low effort. Foundational scenarios often reward choosing the service that solves the business problem without unnecessary custom engineering.
- Prefer controls when the scenario says sensitive, regulated, public-facing, or responsible. Security and governance requirements are not optional details.
- Guess rather than leave blank. AWS states that unanswered questions are scored as incorrect and there is no penalty for guessing.
When You Are Ready To Schedule
Schedule only when you can do these things without notes:
- Explain the five AIF-C01 domains and roughly how heavily they are weighted.
- Tell the difference between AI, ML, deep learning, GenAI, foundation models, LLMs, embeddings, RAG, fine-tuning, and agents.
- Choose between Amazon Bedrock, Amazon Q, SageMaker AI, Comprehend, Rekognition, Transcribe, Translate, Lex, Polly, Kendra, and core security/monitoring services in simple scenarios.
- Identify responsible AI risks such as bias, hallucination, toxicity, privacy exposure, poor transparency, and lack of human oversight.
- Apply IAM, KMS, CloudTrail, CloudWatch, and shared responsibility concepts to AI use cases.
- Complete mixed practice sets at a passing margin, not just one memorized topic set.
Official Links For This Module
- AWS Certified AI Practitioner exam page - cost, duration, question count, candidate profile, languages, and scheduling entry point.
- Official AIF-C01 exam guide - target candidate, question types, scored content, passing score, and domain weights.
- AIF-C01 in-scope AWS services - services and features that can appear in exam scenarios.
- AWS scheduling and testing options - Pearson VUE test center and online proctored exam information.
- AWS before-testing policies - pricing policy, reschedule, cancellation, and appointment rules.
- AWS after-testing policies - score reporting, passing standards, retakes, and beta exam rules.
- AWS recertification page - validity period and recertification options.
Before You Move On
You are ready for the next module when you can answer these questions clearly: Who is the AIF-C01 exam for? What does it cost? What is the retake wait after a failed attempt? What are the five exam domains and weights? Which topics are out of scope? How should you approach multiple response, ordering, and matching questions?
If any answer is vague, review the official exam guide and repeat a small flashcard set before moving into AI and Data Foundations.