Amazon Bedrock Guardrails announces new capabilities to safely build generative AI applications

Amazon Bedrock Guardrails announces new capabilities to safely build generative AI applications at scale. These new capabilities offer greater flexibility, finer-grained control, and ease of use while using the configurable safeguards provided by Bedrock Guardrails aligning with use cases and responsible AI policies.

Bedrock Guardrails now offers a detect mode that provides a preview of the expected results from your configured policies allowing you to evaluate the effectiveness of the safeguards before deploying them. This enables faster iteration and accelerating time-to-product with different combinations and strengths of policies, allowing you to fine-tune your guardrails before deployment.

Guardrails now offers more configurability with options to enable policies on input prompts, model responses, or both prompts and responses - a significant improvement over the previous default setting where policies were automatically applied to both inputs and outputs. Providing finer-grained control enables you to selectively apply the safeguards to make them work for you.

Bedrock Guardrails offers sensitive information filters that detect personally identifiable information (PIIs) with two modes: Block where requests containing sensitive information are completely blocked, and Mask where sensitive information is redacted and replaced with identifier tags. You can now use either of these two modes for both input prompts and model responses giving you flexibility and ease of use to safely build generative AI applications at scale.

These new capabilities are available in all AWS regions where Amazon Bedrock Guardrails is supported.

To learn more, see the blog post, technical documentation and the Bedrock Guardrails product page.