Building secure AI agents at scale: Introducing Loom for AWS

This article introduces Loom for AWS, an enterprise-grade platform for building and deploying AI agents securely at scale with governance and security controls.

• Unified management UI with identity provider integration, role-based and attribute-based access controls, and lifecycle management
• Automated resource tagging for cost attribution and governance compliance
• Paved path blueprints with pre-validated configurations for secure agent deployments
• Configuration-driven deployments using pre-written agents with no runtime code generation
• Identity propagation through delegated actor chains using RFC 8693 token exchange
• Integration with AWS Agent Registry for agent discovery and governance
• Human-in-the-loop approval workflows for sensitive agent actions

Loom provides a simple, opinionated platform that abstracts AWS managed services complexity while enforcing enterprise security best practices for agentic application deployments.