*Note: This course has been updated to enhance your learning experience. The learning objectives remain the same. As part of this update, in-progress status may be reset. If you have already completed this course, no action is needed.
*
In this module, you'll learn to protect Generative AI applications against malicious inputs, inappropriate content generation, data exposure, and prompt injection attacks. You'll implement comprehensive safety strategies including input filtering, model response controls, and real-time validation using Amazon Bedrock Guardrails and AWS WAF. The module covers output safety techniques such as content moderation, toxicity detection with confidence metrics, Text-to-SQL transformations, grounding checks, semantic similarity search for fact-checking, and structured outputs with JSON Schema. You'll gain practical skills to develop secure Generative AI applications that maintain safety and integrity while delivering user value.
