AWS Certified AI Practitioner (AIF-C01) is the entry point into AWS's AI/ML certification track. It isn't a coding exam — it's a Foundational-level credential that tests whether you understand why, when, and how to use AI services, so you don't need to be a developer to take it on.
This post opens an 11-part domain deep-dive series. Before going topic by topic, this guide draws the full map: exam structure, who the exam is actually for, and where AIF-C01 sits in the broader AWS AI/ML career path.
---
Who Is AWS Certified AI Practitioner For?
AIF-C01 doesn't ask how well you can code. It asks whether you understand what AI services can and can't do, and whether you can pick the right one for a given situation. Think of it like learning to drive: you don't need to know how the engine works internally, just when to use which gear and why seatbelts matter.
The certification fits a wide range of roles:
| Target Audience | Why It Fits | |-----------------|-------------| | Non-technical roles (PM, sales, marketing, legal/compliance) | Understand AI capabilities and risk without writing code, to inform business decisions | | Cloud/IT staff new to AWS | The first stop on the AWS AI/ML career path, before MLA-C01 and AIP-C01 | | Data analysts and BI professionals | Extend into Bedrock-powered GenAI analytics tooling | | Career changers moving into AI/ML | A low-risk way to validate foundational knowledge before specializing | | Engineering managers and team leads | Build the judgment needed to evaluate AI service choices and governance for a team |
There are no prerequisites. You can register today with zero AWS experience — though in practice, a few weeks of hands-on time in the console and with Bedrock make the difference between guessing and passing.
---
Exam at a Glance
| Item | Detail | |------|--------| | Exam Code | AIF-C01 | | Full Name | AWS Certified AI Practitioner | | Passing Score | 700 (scale 100–1,000) | | Total Questions | 65 (50 scored + 15 unscored) | | Duration | 90 minutes | | Exam Fee | $100 USD | | Validity | 3 years | | Question Types | Multiple choice, multi-select, matching | | Recommended Experience | 6 months with AWS AI/ML (not required) |
The 15 unscored questions are AWS testing material for future exams, and you won't know which ones they are — so treat every question the same. If you don't pass, you must wait 14 days before retaking, and that same 14-day wait applies to every subsequent attempt.
---
Domain Breakdown
| Domain | Topic | Weight | |--------|-------|--------| | Domain 1 | AI and ML Fundamentals | 20% | | Domain 2 | Generative AI Fundamentals | 24% | | Domain 3 | Applications of Foundation Models | 28% | | Domain 4 | Responsible AI Guidelines | 14% | | Domain 5 | Security, Compliance, Governance | 14% |
Domain 3 has the highest weight (28%) and focuses on Amazon Bedrock, prompt engineering, and RAG.
!AIF-C01 exam domain weight breakdown across 5 domains
Domain 1: AI & ML Basics (20%)
Key topics: Supervised vs. unsupervised vs. reinforcement learning, neural networks, model training fundamentals.
Exam tip: "Fraud detection with labeled historical data" → Supervised learning.
Related posts: Core Concepts of AI and ML, ML Development Lifecycle
Domain 2: Generative AI Basics (24%)
Key topics: Tokens, embeddings, Foundation Models, LLMs, prompt engineering techniques, RAG vs. fine-tuning, hallucinations.
Exam tip: Prompt engineering does NOT update model weights. Fine-tuning does. Know the difference.
Related posts: Generative AI Core Concepts, Generative AI Capabilities and Limitations, Complete Guide to FM Training and Fine-tuning
Domain 3: Foundation Model Applications (28%)
Key topics: Amazon Bedrock, Bedrock Knowledge Bases, Bedrock Agents, inference parameters, SageMaker JumpStart.
Exam tip: Understand the cost-performance tradeoff. Smaller models = lower cost, faster speed. Larger models = higher accuracy, multilingual support.
Related posts: AWS GenAI Infrastructure and Services, FM App Design and RAG, Prompt Engineering and AI Security, AWS AI Use Cases and Service Selection Guide
Domain 4: Responsible AI (14%)
Key topics: Bias types, SageMaker Clarify, Amazon A2I, Bedrock Guardrails.
Related posts: Complete Guide to Responsible AI
Domain 5: Security & Governance (14%)
Key topics: AWS Config, CloudTrail, Model Cards, Shared Responsibility Model, IAM for AI.
Related posts: Complete Guide to AI Transparency and Governance
---
Where AIF-C01 Fits in the AWS AI/ML Certification Path
AWS lays out a three-tier path for AI/ML careers. AIF-C01 is where it starts.
| Attribute | AIF-C01 | MLA-C01 | AIP-C01 | |-----------|---------|---------|---------| | Level | Foundational | Associate | Professional | | Full Name | AI Practitioner | Machine Learning Engineer | Generative AI Developer | | Primary Audience | Any role, including non-technical | Engineers who build and operate ML pipelines | Developers who build and deploy Bedrock-based GenAI apps | | Coding Required | No | Yes, SageMaker hands-on | Yes, Bedrock API/SDK hands-on | | Exam Fee | $100 | $150 | $300 |
The Bedrock, prompt engineering, and Responsible AI concepts you build for AIF-C01 carry directly into both higher-level exams. If you want to grow into an engineer who builds and operates ML systems, continue with the AWS MLA-C01 Complete Exam Guide. If a generative AI application developer career is the goal, move to the AWS AIP-C01 Exam Guide.
---
Study Roadmap for Beginners
Step 1 (Weeks 1–2): Learn AI/ML fundamentals using AWS Skill Builder free courses. No coding needed.
Step 2 (Weeks 3–5): Explore Amazon Bedrock in the AWS console. Try prompting Claude or Titan models directly.
Step 3 (Weeks 6–7): Practice with official AWS Skill Builder questions. Focus on Domains 3 and 4.
Step 4 (Week 8): Review wrong answers. Create a cheat sheet of service names and use cases.
---
Exam Day Checklist
You can sit the exam through Pearson VUE, either via online proctoring or at a physical test center. For online proctoring, log in 30 minutes early — you'll need to show a photo ID, clear your desk, and confirm a single-monitor setup before the session starts.
One government-issued photo ID (passport, driver's license, etc.) No pens, calculators, or personal notes during the exam — only the on-screen notepad For in-person testing, arrive at least 15 minutes early Failed attempts require a 14-day wait before retaking; certification is valid for 3 years
---
Exam Quick Reference
| Topic | Key Point | |-------|-----------| | Hallucination | AI generates plausible but false content | | Prompt Engineering | Adjust inputs only, no model retraining | | Fine-tuning | Updates model weights for domain specialization | | RAG | Retrieves external documents to augment prompts | | Bedrock Guardrails | Bidirectional safety filter including PII masking | | SageMaker Clarify | Bias detection + model explainability | | Amazon A2I | Human review loop for AI decisions | | Shared Responsibility | AWS secures infrastructure; customer secures data and model |