Accelerating government innovation: Amazon Bedrock models get FedRAMP High and DoD IL-4/5 approval in AWS GovCloud (US)

Anthropic’s Claude and Meta’s Llama models within Amazon Bedrock in the AWS GovCloud (US) Regions are now FedRAMP High and DoD IL4/5 approved

In today’s rapidly evolving security landscape, the technological edge isn’t just an advantage—it’s a necessity. I’m thrilled to share a milestone that will transform how our government customers harness the power of artificial intelligence (AI): Amazon Web Services (AWS) is the first cloud provider to achieve FedRAMP High and Department of Defense (DoD) Cloud Computing Security Requirements Guide Impact Level 4 and 5 authorizations for Anthropic’s Claude and Meta’s Llama AI foundation models.

This achievement represents a pivotal moment in public sector innovation by ensuring government agencies have secure, compliant access to AI tools with scalable capabilities and advanced features.

Enhancing America’s competitive edge with secure AI

Our defense customers recognize the importance of accelerating technological adoption to maintain and enhance its competitive advantage in an evolving global landscape—without compromising on security. With the launch of AWS GovCloud (US) in 2011, we became the first cloud provider to build cloud infrastructure specifically designed to meet U.S. government security and compliance needs. This authorization continues to demonstrate our commitment to maintaining a high security bar for our nation’s sensitive data, while enabling leaders to harness the full potential of AI and generative AI in AWS GovCloud (US) Regions.

AWS’s built-in security and compliance framework is key to innovation speed. What makes this authorization particularly powerful is the ability for agencies to develop, test, and deploy generative AI applications within a single, compliant environment—dramatically accelerating the timeline from proof-of-concept to operational advantage. Instead of spending months navigating complex security processes, leaders can focus on what matters: delivering enhanced services and capabilities to support their missions.

Transforming mission capabilities with leading AI models

This authorization delivers unprecedented access to advanced foundation models within security-compliant environments, enabling a range of mission-critical applications. Analysts can now process vast amounts of sensitive data at unprecedented speed, identifying patterns human analysis might miss. Strategic planning is revolutionized through AI-powered tools that optimize strategies by simulating countless scenarios. Decision making that once required thousands of personnel hours can now be automated while maintaining compliance with security protocols. Advanced analytics capabilities give mission owners unprecedented insight into evolving challenges.

“With this achievement, AWS is expanding the potential for Meta’s open source Llama models to power mission-critical applications in secure or disconnected environments at a lower cost. We are proud to support America’s defense agencies with cutting edge AI technology that helps maintain their competitive edge in an increasingly complex global landscape,” said Molly Montgomery, Director of Public Policy at Meta.

“With Amazon Bedrock’s FedRAMP High and DoD IL4/5 authorization, Claude can now assist defense agencies with their most sensitive missions while ensuring compliance with the strictest security requirements,” said Thiyagu Ramasamy, Head of Public Sector at Anthropic. “This authorization opens new possibilities for responsible AI use in scenarios where both performance and security are essential for serving the public interest.”

Building the future of defense with advanced AI features

With the authorization of these models in Amazon Bedrock, defense and public sector organizations gain access to a comprehensive suite of features specifically designed for mission-critical applications. Agencies can now deploy AI agents that execute complex multi-step tasks across security-compliant systems. Knowledge Bases enable foundation models to leverage agency-specific information from private data sources within security boundaries. Custom Guardrails allow teams to implement safeguards aligned with agency-specific responsible AI policies. Knowledge Store securely manages data sources for retrieval augmented generation applications.

These capabilities deliver immediate benefits that address longstanding challenges in defense technology adoption: simplified compliance through a unified API on authorized infrastructure, streamlined Authority to Operate (ATO) processes, and the ability to build and customize models for agency-specific use cases while implementing strict controls for responsible AI use.

As world security challenges grow increasingly complex, the combination of advanced AI capabilities alongside public sector expertise will become a defining factor in maintaining American AI leadership. By putting generative AI capabilities directly into the hands of those serving our nation, we ensure America’s leaders have the tools they need to effectively address evolving challenges in an increasingly complex global environment.

To learn more about using Amazon Bedrock in AWS GovCloud (US), visit the Amazon Bedrock in GovCloud (US) documentation or contact your AWS account team.