AWS Verified Access provides access to corporate applications without requiring users to join a corporate network through a traditional VPN. It places a managed access layer in front of applications and evaluates authorization using configured trust information and access policies. For browser-based applications, engineers can expose an HTTPS application hostname while keeping the application backend private inside an Amazon VPC.
The main configuration objects are Verified Access instances, trust providers, groups, and endpoints. An instance establishes the trust configuration used to evaluate access. Trust providers supply user identity or device posture information. A group collects endpoints that share a baseline policy, while each endpoint identifies an application backend and can have an additional application-specific policy. Access must satisfy the applicable group and endpoint policies.
User identity can come from AWS IAM Identity Center or a supported OpenID Connect identity provider. Device trust providers can supply security signals from supported device-management and endpoint-security systems. Policies use the Cedar policy language and reference attributes made available through the configured trust providers. For example, an engineering application can require membership in a particular identity group and a device compliance signal. Attribute names and available values depend on the provider, so inspect the actual context before writing production policies.
For HTTPS applications, endpoint configuration includes the application domain, a suitable TLS certificate, and connectivity to a supported backend such as a load balancer or network interface. Engineers must configure DNS, subnets, security groups, and backend connectivity correctly. Verified Access performs access enforcement, but it does not repair routing problems or replace application-level authorization. Applications should continue to enforce permissions for individual records, operations, and administrative functions.
AWS Verified Access also supports non-HTTP TCP access through supported endpoint configurations and an AWS-provided connectivity client. This differs from browser-based HTTPS access in deployment requirements, client behavior, and billing. Confirm protocol, client-platform, Region, and trust-provider support before selecting an architecture for database or other TCP-based access.
Operationally, enable access logging and monitor denied requests, policy changes, and backend failures. Access logs can be delivered to supported AWS logging destinations, and AWS CloudTrail records supported management API activity. Treat trust-provider configuration and Cedar policies as security-sensitive infrastructure: version them, review changes, and test both permitted and denied access paths. When migrating from a VPN, start with a limited application set, validate authentication and session behavior, and remove obsolete backend access paths so users cannot bypass the new controls.