Control where your AI agents can browse with Chrome enterprise policies on Amazon Bedrock AgentCore

This article explains how to control AI agent browser behavior using Chrome enterprise policies and custom root CA certificates in Amazon Bedrock AgentCore.

• Chrome enterprise policies restrict agent browsing to approved domains and disable risky features
• Over 450 configurable Chrome settings available including URL filtering and download restrictions
• Custom root CA certificates enable agents to connect to internal services and corporate proxies
• Managed policies applied at browser level; recommended policies at session level
• Policies stored in S3; certificates stored in AWS Secrets Manager
• Session recording allows replay of policy enforcement in action
• Works with both AgentCore Browser and Code Interpreter tools
• Complete walkthrough provided with sample Jupyter notebook and code examples

Organizations can now deploy AI agents with granular security controls, restricting access to approved resources while maintaining connectivity to internal infrastructure through certificate trust configuration.