Amazon Bedrock AgentCore Identity
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 4
- AgentCore Identity adds private CA support
Amazon Bedrock AgentCore Identity now supports private CA certificates for connecting agents to self-hosted identity providers. According to AWS Machine Learning Blog, users can store a PEM-encoded root or intermediate CA certificate in AWS Secrets Manager or Amazon S3, then reference it from a credential provider for endpoint validation, with VPC Lattice required for private connectivity.
AWS Machine Learning Blog · 🔥 0 - AWS details OAuth patterns for Snowflake agents via AgentCore
The AWS Machine Learning Blog details three OAuth configurations tested end-to-end for connecting AI agents to Snowflake’s managed MCP server through Amazon Bedrock AgentCore Gateway: Snowflake OAuth, three-legged OAuth with an external identity provider, and two-legged OAuth using a shared service user. AWS says AgentCore Identity manages credential brokering, token storage, refresh, and user-to-token binding, while also claiming support for an on-behalf-of TOKEN_EXCHANGE grant for unattended workloads. The gateway exposes Snowflake Cortex Search and Cortex Analyst through a unified MCP tool catalog.
AWS Machine Learning Blog · 🔥 0 - AWS tutorial scales governed AI agent access through AgentCore Gateway
AWS Machine Learning Blog published Part 2 of a tutorial showing how an AI agent on Amazon Bedrock AgentCore can act for signed-in users through Okta Cross App Access. The implementation routes agents through AgentCore Gateway, evaluates a default-deny Cedar policy set before tool execution, and connects governance decisions to Okta logs through AgentCore Observability and Amazon CloudWatch Omni.
AWS Machine Learning Blog · 🔥 0 - AWS Details Governed AI Agent Access With AgentCore and Okta
AWS Machine Learning Blog published a Part 1 guide to governing AI agent access with Okta Cross App Access and Amazon Bedrock AgentCore. It describes an architecture in which Okta authorizes each user-agent pairing, while AgentCore validates the grant and uses its token vault to obtain short-lived credentials without exposing them to agent code. AWS reported that revoked assignments caused new requests to be refused in about one second, existing tokens expired within five minutes, and AgentCore Identity added about 0.6 seconds to a roughly 30-second request.
AWS Machine Learning Blog · 🔥 0
Experience and discussion from the community
Share my Amazon Bedrock AgentCore Identity experienceAsk about Amazon Bedrock AgentCore Identity
Nobody has shared their experience with Amazon Bedrock AgentCore Identity yet.