AWS Certified AI Practitioner: Skills Measured

The official AIF-C01 exam guide lists five domains: Fundamentals of AI and ML, Fundamentals of Generative AI, Applications of Foundation Models, Guidelines for Responsible AI, and Security, Compliance, and Governance for AI Solutions. Treat the domains as connected. A foundation-model scenario may also test prompt design, grounding, evaluation, data privacy, and human review.

This domain checks whether you understand AI, ML, deep learning, neural networks, computer vision, NLP, training, inference, labeled and unlabeled data, and the main learning types. You should also know when regression, classification, clustering, recommendation, forecasting, or knowledge retrieval is an appropriate pattern and when AI is not the right solution.

Foundation-model questions focus on applying models to real use cases. Know how prompts, context, embeddings, knowledge bases, retrieval, grounding, evaluation, and human review fit together. Amazon Bedrock is central here because it provides access to foundation models and supporting features for building generative AI applications.

Generative AI questions test how models produce content and how output can be influenced or constrained. Be ready for tokens, context windows, temperature, hallucinations, prompt engineering, RAG, model selection, and the difference between grounding, fine-tuning, and ordinary prompt changes. The exam expects concept recognition, not deep model math.

Responsible AI covers fairness, bias, explainability, transparency, safety, privacy, and human-centered design. Governance and security questions bring in access control, auditability, encryption, data protection, CloudTrail, IAM, AWS KMS, Secrets Manager, and monitoring. The recurring skill is knowing which control reduces which AI risk.