AI Sec Watch: A Security Intelligence Platform for AI Systems

Luu, T.J.

Secure multi-tenant AI agents with Amazon Bedrock AgentCore resource-based policies

infonews

securitypolicy

Source: AWS Security BlogJune 2, 2026

Summary

Amazon Bedrock AgentCore is a tool that lets Software as a Service (SaaS) providers serve multiple clients, called tenants, with different security needs using the same AI agent. Resource-based policies (rules that control who can access a resource directly) let you grant some tenants cross-account access from their own AWS accounts while restricting others to traffic that stays only within a private virtual network, all without sharing credentials or creating separate user accounts for each tenant.

Solution / Mitigation

Use resource-based policies on AgentCore Runtime and AgentCore Runtime endpoint resources to centralize access control. For cross-account access (like Example Corp), implement both a resource-based policy on your resources and an identity-based policy (access rules tied to a user or role) in the tenant's AWS account. For VPC-restricted scenarios (like AnyCompany), use specific IAM conditions to enforce that requests originate only from an approved virtual private cloud (VPC, a private network in AWS), adding a network-level security boundary on top of identity-based controls.