AIP-C01 25/36: Input and Output Safety Controls | AWS Certified Generative AI Developer Professional

Input and Output Safety Controls explained for the AWS Certified Generative AI Developer Professional exam: protect inputs and outputs with Amazon Bedrock Guardrails (content filters including prompt attack, denied topics, word filters, PII filters, grounding and Automated Reasoning checks), apply them selectively with input tags or Converse guardrail content blocks, use ApplyGuardrail for models outside Bedrock, add custom moderation with Step Functions and Lambda, and use text-to-SQL for deterministic answers. Input and Output Safety Controls is in Domain 3, AI Safety, Security, and Governance, which is 20% of the AIP-C01 exam. This exam-focused lesson is for developers building production generative AI applications on AWS preparing for the certification. What you will learn: • Guardrails policies, revisited for builders • How a guardrail evaluates a call • Evaluating only the user input • ApplyGuardrail for any model • Custom moderation and deterministic answers • Exam traps and self-check questions CHAPTERS 00:00 Input and output safety controls for the AIP-C01 exam 00:32 Guardrails policies, revisited for builders 01:08 How a guardrail evaluates a call 01:38 Evaluating only the user input 02:12 ApplyGuardrail for any model 02:38 Custom moderation and deterministic answers 03:07 Exam traps for safety controls 03:35 Self-check questions AIP-C01 full course: video 25 of 36, in blueprint order (Domain 1 to Domain 5). Next: Hallucination Controls and Defense in Depth. Facts checked against the official AWS AIP-C01 exam guide and AWS documentation on 2026-10-19. Exam format and domain weights are from the official AWS exam guide for AIP-C01 (checked October 2026). Unofficial study material, not affiliated with or endorsed by Amazon Web Services. Confirm against the current exam guide before you sit the exam. Subscribe to Core Concept Learning for clear, visual explanations of AI and software engineering concepts. #AIPC01 #AWSCertified #ResponsibleAI