Back to Blog
Architecture

AWS Bedrock AgentCore Reference Architecture for Safe Agent Deployment

Five-layer blueprint — IAM scope, Guardrails, Action Groups, CloudTrail — that passes a financial-services security review

AWS Bedrock AgentCore handles compute and orchestration. It does not handle the governance layer. Here is the five-layer reference architecture — IAM scoping, Guardrails, Action Group contracts, runtime policy, and CloudTrail — that passes a financial-services cloud security review.

July 29, 202612 min read
AWS BedrockAgentCoreReference ArchitectureIAMGuardrailsSOC 2Financial Services
AgentTrustOSAGENTIC AI GOVERNANCEAWS · REFERENCE ARCHITECTUREAWS Bedrock AgentCore:Reference Architecture forSafe Agent DeploymentAWS BEDROCK AGENTCORE STACKApplication LayerAgentCore RuntimeCedar Policy EnforcementTool Execution + IAM IdentityAudit Lake + CloudWatchSources: AWS Bedrock AgentCore GA (Oct 2025) · Cedar Policy · AWS Security Hub · AWS CloudWatchagent-trust.tech
AWS Bedrock AgentCore five-layer stack: Application → AgentCore Runtime → Cedar Policy → Tool Execution → Audit Lake
AWS Architecture · Bedrock AgentCore · Cloud Security
Key Facts
  • AWS Bedrock AgentCore Runtime reached general availability in October 2025, providing managed infrastructure for agent lifecycle management, tool execution, and session state without requiring custom orchestration framework development.
  • AWS AgentCore Policy, powered by Amazon Cedar's attribute-based access control engine, reached general availability in March 2026, providing an enforcement point for every tool call an agent attempts — before execution.
  • AWS Bedrock Guardrails provides model-layer protection including PII detection and redaction, grounding checks to prevent hallucinated content from propagating, and content filtering for prompt injection patterns — complementing AgentCore's tool-layer enforcement.
  • IAM plus AgentCore Identity enables per-agent credential issuance, assigning each agent a distinct, task-scoped credential that eliminates the shared-secret anti-pattern that violates OWASP LLM06 least-privilege requirements.
  • The AWS audit lake pattern — CloudWatch and X-Ray for real-time traces, plus S3, CloudTrail, and AWS Glue for immutable long-term storage and analysis — provides the compliance-grade audit record required by SOC 2, ISO 27001, and emerging AI governance frameworks.
  • AWS Bedrock Knowledge Bases provides managed RAG (Retrieval-Augmented Generation) infrastructure, eliminating the need to manage separate vector database infrastructure for agents that require grounded responses from organizational knowledge.
TL;DR
  • AWS Bedrock AgentCore Runtime (GA Oct 2025) + AgentCore Policy (GA Mar 2026) give platform teams a complete governed agent stack on AWS — no custom orchestration required.
  • The enforcement layer is AgentCore Gateway + Cedar Policy: every tool call is gated by Cedar's attribute-based policy engine before it executes.
  • IAM + AgentCore Identity provides per-agent credentials, eliminating shared secrets; Bedrock Guardrails provides model-layer PII and grounding protection.
  • The audit lake (CloudWatch/X-Ray + S3/CloudTrail/Glue) gives compliance teams the forensic record they need without custom logging infrastructure.
  • This is the AWS-native equivalent of the Azure APIM pattern — different managed services, same governance outcome.
Keep reading → Full layer-by-layer architecture breakdown below.

Platform teams at AWS-native financial institutions have been watching the agentic AI space and asking the same question: "When the managed services are ready, can we build a governed agent stack without custom middleware?" As of early 2026, the answer is yes. AWS Bedrock AgentCore Runtime, which went GA in October 2025, and AWS AgentCore Policy, which followed in March 2026, together provide the enforcement, identity, and audit infrastructure that governed agent deployment requires.

The challenge before these services was real: organizations that wanted to deploy AI agents in production needed to either use a third-party orchestration framework (LangChain, CrewAI, AutoGen) with custom governance wrapping, or build their own agent runtime infrastructure from scratch. Neither option was operationally attractive for regulated industries where the governance overhead of custom software is itself a compliance burden. AWS AgentCore changes the calculus by providing managed infrastructure with governance built in.

What is AWS Bedrock AgentCore and what problem does it solve?

AWS Bedrock AgentCore is a managed service that provides the runtime infrastructure for deploying, executing, and governing AI agents built on foundation models available in Amazon Bedrock. It addresses the operational gap that existed between Bedrock's model invocation APIs — which allow you to call a foundation model — and the production requirements of an agent that needs to maintain session state, invoke external tools, track execution history, and respect governance policies.

Definition — AWS Bedrock AgentCore Runtime

AWS Bedrock AgentCore Runtime, generally available as of October 2025, is the managed execution environment for AI agents on AWS. It provides: agent lifecycle management (creation, versioning, deployment, and deprecation), tool execution infrastructure (managed invocation of tools defined in the agent's tool configuration), session state management (maintaining conversation context across turns), and integration with AWS governance services including IAM, CloudWatch, and the AgentCore Gateway/Cedar Policy enforcement layer.

How does AgentCore Gateway and Cedar Policy provide enforcement?

AWS AgentCore Gateway is the traffic management layer that sits between an agent's intention (the tool call it wants to make) and the actual execution of that tool call. Every tool call an agent attempts passes through the Gateway, where it is evaluated against Cedar policies before execution proceeds. Cedar is Amazon's open-source attribute-based access control (ABAC) language — it defines policies in terms of the agent's identity, the requested action, the target resource, and the context of the request.

Definition — Amazon Cedar Policy Language

Cedar is an open-source policy language developed by AWS for expressing attribute-based access control (ABAC) decisions. Cedar policies specify conditions under which a principal (such as an AgentCore agent) may perform an action (such as invoking a tool) on a resource (such as a payment API), given contextual attributes (such as transaction amount, customer tier, or time of day). Cedar is designed for correctness and auditability: policies are formally verifiable and the decision reasoning for each authorization is available for logging. AWS AgentCore Policy uses Cedar to gate every tool call in the AgentCore Runtime.

The governance value of Cedar Policy enforcement is that it is preventive, not detective. Traditional security monitoring detects anomalies after an agent has already taken an action. Cedar Policy enforcement prevents the action from executing in the first place if it violates policy. For a financial services organization where an incorrect payment or unauthorized data access cannot be easily reversed, the distinction between "we detected it" and "we prevented it" is significant.

In practice, Cedar policies for agent governance express rules like: "Agent payment-processor-v2 may invoke the refund API only for amounts less than $500 and only during business hours" or "Agent customer-data-agent may read customer records but may not modify or delete them." These policies are managed centrally, versioned, and audited independently of the agent's model — meaning a policy change doesn't require redeployment of the agent, and a model update doesn't inadvertently change the policy.

What does Bedrock Guardrails add that Cedar Policy doesn't cover?

Bedrock Guardrails and Cedar Policy operate at different layers of the agent stack and are complementary rather than redundant. Cedar Policy operates at the tool call layer — it governs what tools the agent may invoke and under what conditions. Bedrock Guardrails operates at the model input/output layer — it governs what content the agent may receive from users and what content it may output, regardless of which tool is eventually called.

Bedrock Guardrails provides three protection capabilities directly relevant to regulated industry deployments: PII detection and redaction automatically identifies and removes personal information from model inputs and outputs; grounding checks evaluate whether the agent's response is factually consistent with the retrieved context from Knowledge Bases; and content filtering blocks prompt injection patterns before they reach the model's context window.

How do IAM and AgentCore Identity provide per-agent credentials?

AWS AgentCore Identity, integrated with AWS Identity and Access Management (IAM), provides the per-agent credential infrastructure that addresses OWASP LLM06 Excessive Agency's requirement for per-agent identity rather than shared credentials. Each agent deployed in AgentCore Runtime receives a distinct IAM identity — an agent-specific role with a policy that maps to its Cedar Policy-defined tool access permissions. The agent's credential is session-scoped and time-limited, automatically expiring when the session ends.

This eliminates the shared-secret anti-pattern where all agents share a single service account or API key. With per-agent identity, if a credential is compromised, the blast radius is limited to the specific agent's session and the specific tools that agent was authorized to invoke. The audit trail attributes every action to the specific agent identity rather than to an anonymous shared account.

Oct 2025
AWS Bedrock AgentCore Runtime GA date
AWS, AgentCore GA announcement
Mar 2026
AWS AgentCore Policy (Cedar) GA date
AWS, AgentCore Policy GA announcement
5
Governance layers in the full AgentCore reference stack
AgentTrust OS architecture analysis, 2026

How do you build an audit lake for agent governance on AWS?

The AWS audit lake for agent governance combines three service categories: real-time observability, structured event collection, and long-term immutable storage with query capability. Together they provide the compliance record that regulated industries require — a comprehensive, tamper-resistant log of every agent action, decision, tool call, and governance event.

Real-time observability is provided by Amazon CloudWatch and AWS X-Ray. CloudWatch captures structured logs from AgentCore Runtime, including agent session events, tool call attempts, policy evaluation results, and errors. X-Ray provides distributed tracing — for each agent session, X-Ray generates a trace that shows the complete execution path from the initial user request through every tool call, policy check, and model invocation.

Structured event collection is handled by AWS CloudTrail, which captures API-level events from every AWS service the agent interacts with. CloudTrail logs are append-only by default and can be configured to write to a dedicated S3 bucket with Object Lock enabled, making them tamper-resistant. Long-term storage and analysis is provided by S3 (for raw log storage) and AWS Glue (for cataloging and transforming logs for analytical query with Amazon Athena).

AWS Bedrock AgentCore — Reference Architecture for Governed Agent DeploymentApplication / User RequestAPI Gateway · SDK · CLIBEDROCK AGENTCORE RUNTIMEAgent lifecycle · Tool execution · Session state · Version managementAGENTCORE GATEWAY + CEDAR POLICYEvery tool call gated · ABAC enforcement · Policy versioning · Decision auditBEDROCKGUARDRAILSPII · Grounding · InjectionTOOL EXECUTION LAYERExternal APIs · Bedrock KB (RAG) · Lambda · Step FunctionsIAM + AGENTCOREIDENTITYPer-agent credentialsCLOUDWATCH+ X-RAYReal-time tracesAUDIT LAKES3 (immutable storage, Object Lock) · CloudTrail (API events) · AWS Glue (catalog + ETL)Amazon Athena (SQL query) · Compliance export (SOC 2 · ISO 27001 · AI governance)BEDROCK KNOWLEDGE BASESManaged RAG · Semantic search · Grounded responses · No external vector store required
Figure 1: AWS Bedrock AgentCore Reference Architecture. Layers from top: Application → AgentCore Runtime → Cedar Policy Enforcement → Tool Execution (+ IAM Identity + CloudWatch) → Audit Lake → Knowledge Bases. Bedrock Guardrails applies at the model layer alongside enforcement.

How does this architecture compare to the Azure APIM enforcement pattern?

The Azure reference architecture for governed agent deployment uses Azure API Management (APIM) as the enforcement gateway, Microsoft Entra Agent ID for per-agent identity, and Azure Monitor with Log Analytics for audit observability. The AWS architecture uses AgentCore Gateway + Cedar Policy for enforcement, IAM + AgentCore Identity for per-agent credentials, and CloudWatch/X-Ray + CloudTrail for observability.

Both architectures solve the same governance problem — every tool call gated, per-agent identity, immutable audit record — using the managed services available in their respective clouds. The key difference is the policy language: Azure APIM uses inbound/outbound policy XML; AWS Cedar uses a dedicated, formally verifiable ABAC language. Cedar's formal verification properties — policies can be mathematically proven to be correct before deployment — provide a stronger governance guarantee for regulated industries where policy correctness is a compliance requirement.

Definition — Attribute-Based Access Control (ABAC) for Agents

ABAC is an authorization model that evaluates access decisions based on the attributes of the principal (e.g., agent identity, tier, department), the action requested (e.g., read, write, delete), the resource targeted (e.g., payment API, customer database), and the context of the request (e.g., transaction amount, time of day, customer risk score). For agentic AI, ABAC via Cedar Policy allows governance teams to express complex, context-sensitive policies — "agent X may invoke the refund API for amounts up to $500 during business hours for customers with a risk score below 70" — that cannot be expressed in simple role-based access control (RBAC) models.

AgentTrust OS Integrates Natively with the AWS AgentCore Stack

AgentTrust OS is designed to complement AWS's native governance services, not replace them. For organizations building on the AWS Bedrock AgentCore reference architecture, AgentTrust OS adds the governance layer above the infrastructure — classification, certification, and cross-environment audit aggregation.

Trust Certify
Pre-production certification that validates your Cedar Policy configuration, tool allowlists, and AgentCore Identity bindings against governance policy before agents reach production.
Trust Runtime
Governance policy layer that works alongside AgentCore Gateway and Cedar Policy, adding cross-environment escalation thresholds and human-in-the-loop routing for actions that exceed configured risk thresholds.
Trust Audit
Aggregates CloudWatch, X-Ray, and CloudTrail data from the AWS audit lake into a unified governance reporting layer — generating board-ready reports and regulator-facing compliance exports.

Frequently Asked Questions

Ready to Govern Your Agents?

Deploy on AWS. Govern with AgentTrust OS.

AgentTrust OS integrates natively with AWS Bedrock AgentCore — adding governance classification, certification, and board-ready reporting above AWS's native infrastructure.

Start Free →

More from the blog

AI ComplianceJuly 22, 2026AI ComplianceJuly 22, 2026AI GovernanceJuly 22, 2026AI GovernanceJuly 22, 2026AI ArchitectureJuly 22, 2026AI SecurityJuly 16, 2026MLOpsJuly 10, 2026AI ImplementationJuly 8, 2026AI Agent ArchitectureJuly 5, 2026AI Agent ArchitectureJune 30, 2026EngineeringJune 23, 2026AI Agent ArchitectureJuly 2, 2026AI Agent ArchitectureJuly 1, 2026IntegrationsJuly 1, 2026IntegrationsJuly 1, 2026IntegrationsJuly 2, 2026AI SecurityJuly 3, 2026AI SecurityJuly 1, 2026AI ComplianceJuly 2, 2026AI ComplianceJuly 3, 2026AI StrategyJuly 2, 2026AI StrategyJuly 3, 2026AI StrategyJuly 3, 2026AI StrategyJuly 3, 2026AI GovernanceJuly 28, 2026Healthcare AIJuly 28, 2026ArchitectureJuly 29, 2026AI StrategyJuly 29, 2026AI StrategyJuly 30, 2026AI SecurityJuly 30, 2026AI ComplianceJuly 30, 2026EngineeringJuly 30, 2026AI GovernanceAugust 4, 2026EngineeringAugust 4, 2026EngineeringAugust 4, 2026