Amazon Web Services (AWS) has unveiled Loom for AWS, an open-source platform designed to help organisations build, deploy and govern artificial intelligence (AI) agents securely at scale.
The platform integrates with Amazon Bedrock AgentCore and AWS Strands Agents, providing enterprises with lifecycle management capabilities alongside security controls and governance frameworks for AI agent deployments.
According to AWS, Loom has been developed to address key enterprise requirements by incorporating automated resource tagging, role-based and attribute-based access controls, and pre-validated configuration blueprints for deploying AI agents.
The platform enforces three mandatory tags on all deployed resources while allowing organisations to create additional custom tags for cost allocation and governance. It also introduces a two-dimensional access control model that combines user roles with group tags to ensure permissions are restricted appropriately.
Unlike systems that rely on runtime code generation, Loom deploys AI agents using configuration-driven methods. Behavioural guidelines and security credentials are managed through AWS Secrets Manager, enabling organisations to standardise deployments while strengthening security.
AWS said the platform supports both low-code deployments through pre-written Python agents and no-code deployments using AgentCore’s managed harness. It also incorporates OAuth2 configurations and token exchange protocols, allowing user identities to be securely propagated throughout AI agent request chains.
Loom further integrates with the AWS Agent Registry, which is currently available in public preview, to manage agent and tool records while ensuring governance review processes are completed before production deployment.
To enhance oversight of high-risk operations, the platform includes human-in-the-loop approval workflows powered by the Strands Agents hook framework and Model Context Protocol elicitations, enabling sensitive actions to undergo manual review before execution.
AWS has made the project available through AWS Labs on GitHub and is inviting contributions from the open-source community. The company said Loom is intended primarily for platform engineering teams developing AI-powered applications using fully managed AWS services.
The launch reflects growing demand among enterprises for governance-focused AI infrastructure as organisations seek to deploy autonomous AI agents without compromising security, compliance or operational control.
