Explanation

Guardrails for Amazon Bedrock is the correct answer because it is specifically designed to implement safety controls and content filters for AI applications. Here's why:

Key Features of Guardrails for Amazon Bedrock:

1. Content Filtering: Blocks or filters out harmful, biased, or non-compliant content
2. Topic Control: Allows organizations to define acceptable and prohibited topics
3. Tone Setting: Helps maintain appropriate conversational tone for the application
4. Safety Limits: Implements guardrails to prevent the generation of inappropriate or non-compliant responses

Why Other Options Are Incorrect:

• A) Knowledge Bases for Amazon Bedrock: This feature is for connecting foundation models to company data sources for retrieval-augmented generation (RAG), not for setting safety limits.
• C) Agents for Bedrock: This enables AI agents to perform multi-step tasks using tools and APIs, not for content safety controls.
• D) Continued pre-training: This refers to further training foundation models on specific datasets, which doesn't directly address real-time safety filtering.

Banking Context:

For a bank deploying a chatbot, compliance and bias prevention are critical requirements. Guardrails for Amazon Bedrock provides the necessary controls to ensure the chatbot operates within regulatory boundaries and maintains appropriate professional standards.