Amazon Bedrock Guardrails supports cross-account safeguards with centralized control and management

Amazon Bedrock Guardrails, a service designed to enforce safeguards and policies across AWS accounts, has recently introduced cross-account safeguards. This new feature, now generally available, allows organizations to centralize the enforcement and management of safety controls across multiple AWS accounts within an AWS Organization. This development offers significant benefits, including uniform protection for all accounts and generative AI applications, while providing centralized control and management.

The introduction of cross-account safeguards in Amazon Bedrock Guardrails enables organizations to define a guardrail in a new Amazon Bedrock policy within the management account of their organization. This policy automatically enforces configured safeguards across all member entities for every model invocation with Amazon Bedrock. This organization-wide implementation ensures that all accounts and generative AI applications are protected under a single, unified approach.

One of the key advantages of this new capability is the flexibility it offers in applying account-level and application-specific controls. Organizations can tailor their safeguards to meet specific use case requirements, ensuring that they can enforce policies that are both comprehensive and adaptable.

Organization-level enforcements allow a single guardrail from the organization’s management account to be applied to all entities within the organization through policy settings. This guardrail automatically enforces filters across all member entities, including organizational units (OUs) and individual accounts, for all Amazon Bedrock model invocations. This ensures that the organization-wide safeguards are consistently applied, reducing the administrative burden of monitoring individual accounts and applications.

In addition to organization-level enforcement, account-level enforcement enables automatic enforcement of configured safeguards across all Amazon Bedrock model invocations in a specific AWS account. The configured safeguards in the account-level guardrail apply to all inference API calls, providing an additional layer of protection.

The centralized enforcement in Amazon Bedrock Guardrails supports consistent adherence to corporate responsible AI requirements. By implementing a single, unified approach, organizations can ensure that their AI applications are protected while significantly reducing the administrative burden of monitoring individual accounts and applications. This means that security teams no longer need to oversee and verify configurations or compliance for each account independently, streamlining the overall management process.

To get started with centralized enforcement in Amazon Bedrock Guardrails, organizations can leverage both account-level and organization-level enforcement capabilities. This allows them to establish and centrally manage dependable, comprehensive protection for their AWS accounts and generative AI applications.

In conclusion, the general availability of cross-account safeguards in Amazon Bedrock Guardrails represents a significant step forward in the centralized enforcement and management of safety controls across multiple AWS accounts within an organization. This capability not only ensures uniform protection but also provides flexibility and centralized control, reducing administrative burdens and supporting consistent adherence to corporate responsible AI requirements. As organizations continue to leverage generative AI and other advanced technologies, Amazon Bedrock Guardrails offers a robust solution for safeguarding their AWS environments.