AWS · AIP-C01
Content safety controls
AIP-C01 · Lesson 14 of 23 · 1:52:29
YouTube loads only after you press play. This chapter opens at its timestamp in the full course. Watch on YouTube ↗
In this lesson
Prepare for the AIP-C01 exam (AWS Certified Generative AI Developer - Professional). In this lesson: Bedrock Guardrails content filters, denied topics, word filters, and the prompt attack filter, custom moderation with Step Functions and Lambda, the ApplyGuardrail API, output safety and text-to-SQL, defense in depth with Comprehend, guardrails, Lambda, and API Gateway, and guardrail tiers. Includes three original practice questions.
Preview the topics
- AnyCompany’s assistant must refuse to give investment advice, whatever words the customer uses to ask for it. Which guardrail policy fits best?
- AnyCompany’s agent passes order data from a tool back to the model, and the team assumes that the guardrail’s content filters already screen that tool output. What is the problem?
- AnyCompany wants to screen customer messages for toxicity and unsafe intent before they reach any model, including a model hosted outside Bedrock. Which service can screen them, as a preprocessing filter?