Technologies and concepts that might appear on the exam
The following list contains technologies and concepts that might appear on the exam. This list is non-exhaustive and is subject to change. The order and placement of the items in this list is no indication of their relative weight or importance on the exam:
Core differences between AI, machine learning (ML), and generative AI (GenAI)
How AI models are trained, make predictions, and improve over time
Types of data (for example, structured, unstructured) and why data quality matters for AI
Basics of prompt engineering and how to get useful AI outputs
How context windows and tokens affect GenAI performance
Techniques for improving AI responses (for example, Retrieval Augmented Generation [RAG], fine-tuning)
What AI agents are and how they differ from other AI tools
When to use rule-based automation instead of AI
Why AI systems need ongoing monitoring (for example, model drift, performance changes)
Risks of unapproved AI tool usage (shadow AI) and how to manage risk
Measuring AI value by using key performance indicators (KPIs), return on investment (ROI), cost savings, and productivity gains
Build-buy-partner decisions for AI solutions
Prioritizing AI initiatives by business impact and feasibility
Responsible AI principles (for example, fairness, explainability, privacy, safety, transparency)
Governance structures and who is accountable for AI decisions
Risk classification for AI systems
Regulatory and compliance considerations for AI
Human oversight, guardrails, and knowing when to escalate
Common AI reliability issues (for example, hallucinations, bias, data quality degradation)
Intellectual property (IP) considerations when using AI
Assessing organizational AI readiness and maturity
Data strategy, data ownership, and breaking down data silos
Leading change management and workforce transformation for AI
Building AI literacy across an organization
Scaling AI from pilots to enterprise-wide adoption
AI centers of excellence (COEs)
Amazon Bedrock, Amazon SageMaker AI, and Amazon Quick (at a strategic level)
AWS Cloud Adoption Framework (AWS CAF)
AWS shared responsibility model for AI workloads
AWS AI pricing and cost optimization strategies
Global AI standards and frameworks (for example, ISO/IEC 23053, ISO/IEC 42001)