AWS Certified AI Practitioner Exam Blueprint

Quick Facts

Cost

$100 USD (Foundational tier)

Duration

90 minutes

Questions

65 total (50 scored + 15 unscored)

Passing score

700 / 1000 (scaled, compensatory)

Languages

Arabic, English, French, German, Italian, Japanese, Korean, Portuguese (BR), Spanish (LATAM/ES), Simplified Chinese, Traditional Chinese

Delivery

Pearson VUE test center OR OnVUE online proctored

Validity

3 years

Retake wait

14 calendar days after a fail

Skills You'll Learn

Explain AI, machine learning, deep learning, and generative AI and how they differ. Identify when AI/ML is the right tool for a business problem. Describe the ML development lifecycle from data prep through monitoring. Understand foundation models, tokens, embeddings, vectors, and prompt engineering. Apply prompt engineering - chain-of-thought, few-shot, templates - and avoid common risks. Use Amazon Bedrock, SageMaker, Q, PartyRock, and foundation models to build GenAI solutions. Implement RAG with Bedrock Knowledge Bases and vector stores. Apply Responsible AI - bias detection, fairness, transparency, Bedrock Guardrails, SageMaker Clarify. Secure AI workloads using IAM, encryption, Macie, PrivateLink, AWS GenAI Security Scoping Matrix. Earn AWS's foundational AI credential designed for non-builders who consume AI/ML on AWS.

Exam Logistics

Question types: multiple choice, multiple response, ordering, matching.
Local-currency equivalents shown at checkout (EUR 92, JPY 15,000, etc.).
OnVUE English proctoring 24x7; other languages limited windows.
Recertify by passing the latest AIF-C01 OR earning AWS ML Engineer Associate.

Prerequisites & Recommended Experience

Up to 6 months of exposure to AI/ML on AWS.
No prerequisites - Foundational tier.
Target: uses but does not necessarily build AI/ML - analysts, IT support, marketing, PM, sales.
Out of scope: developing/coding AI/ML models, data engineering, hyperparameter tuning.

Exam Domains with Weights

Domain 1: Fundamentals of AI and ML 20%

AI concepts (AI/ML/deep learning, NLP, LLMs, bias, fairness, supervised/unsupervised/reinforcement).
Practical use cases (regression, classification, clustering; SageMaker, Transcribe, Translate, Comprehend).
ML lifecycle (data collection, EDA, preprocessing, training, deployment, monitoring; SageMaker, Data Wrangler, Feature Store).

Domain 2: Fundamentals of Generative AI 24%

GenAI concepts (tokens, embeddings, vectors, transformers, foundation models, multimodal, diffusion).
Capabilities and limitations of GenAI (hallucinations, interpretability, nondeterminism).
AWS infrastructure (SageMaker JumpStart, Bedrock, PartyRock, Amazon Q, Bedrock Data Automation).

Domain 3: Applications of Foundation Models 28%

Design considerations (cost, modality, latency, customization, Bedrock Agents).
Prompt engineering (chain-of-thought, few-shot, templates; risks - injection, jailbreaking).
FM customization and RAG (Bedrock Knowledge Bases, vector storage - OpenSearch, Aurora).

Domain 4: Guidelines for Responsible AI 14%

Responsible AI features (bias, fairness, safety, veracity; Bedrock Guardrails, SageMaker Clarify, Amazon A2I).
Transparency and explainable models (SageMaker Model Cards).

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

Security (IAM, encryption, Macie, PrivateLink, shared responsibility, prompt injection).
Governance/compliance (Config, Inspector, Audit Manager, CloudTrail, Trusted Advisor; GenAI Security Scoping Matrix).

Official Prep Resources

Exam Prep Plan on AWS Skill Builder Official Exam Guide (HTML) Official Practice Question Set (free) Instructor-led Exam Prep classroom In-scope AWS services list

Test what you've learned

Take a free GoLearnQuiz practice test. Sign in to save your score.

Practice Quiz 60 questions · Multiple choice · Timed

Start now

Additional Helpful Details

Reschedule allowed up to 24 hours before appointment; max 2 reschedules.
Earning AWS ML Engineer Associate auto-recertifies AIF-C01.

AWS Certified AI Practitioner