AWS AI Practitioner (AIF-C01) trap evaluation

Capability boundary confusion and responsible AI misreads

AWS AI services (Rekognition, Comprehend, Transcribe, Translate, Polly, Kendra) and Amazon Bedrock foundation models handle overlapping use cases at different levels of abstraction. The exam tests whether you can identify when a specialized service is the correct fit versus when a foundation model prompt accomplishes the same result with more flexibility but more responsibility. A consistent second trap is responsible AI application: questions present a scenario with a safety or fairness consideration and ask which principle or control applies. Candidates who have studied service capabilities without studying the responsible AI framework misread these as service-selection problems. Reading for whether the scenario is asking about capability or governance is the first filter on each question.