AWS Certified Cloud Practitioner

Explanation

Negative prompting and content filtering is the correct strategy because:

• Negative prompting allows the model to explicitly avoid generating specific content (like competitor brand names)

• Content filtering can be applied to detect and block offensive language or unwanted brand mentions

• This approach directly addresses the requirement to exclude specific content types

Why other options are incorrect:

• A) Zero-shot prompting: This refers to providing examples without training, but doesn't specifically address content exclusion

• C) Reinforcement learning: This is a training methodology, not a content filtering strategy

• D) Hyperparameter tuning: This optimizes model performance but doesn't directly control content generation

Amazon Bedrock provides built-in content filtering capabilities and supports negative prompting techniques to help ensure generated content meets safety and compliance requirements.