GHSA-jp7h-4f3c-9rc7 OpenBao AWS Plugin Vulnerable to Cross-Account IAM Role Impersonation in AWS Auth Method

• Go github.com/openbao/openbao-plugins: < 0.1.1 → fixed in 0.1.1

Impact This is a cross-account impersonation vulnerability in the `auth-aws` plugin. The vulnerability allows an IAM role from an untrusted AWS account to authenticate by impersonating a role with the same name in a trusted account, leading to unauthorized access. This impacts all users of the `auth-aws` plugin who operate in a multi-account AWS environment where IAM role names may not be unique across accounts. The core of the vulnerability is a flawed caching mechanism that fails to validate the AWS Account ID during authentication. While the use of wildcards in a `bound_iam_principal_arn configuration` significantly increases the attack surface, wildcards are not a prerequisite for exploitation. The vulnerability can be exploited with specific ARN bindings if a role name collision occurs. Successful exploitation can lead to unauthorized access to secrets, data exfiltration, and privilege escalation. Given that the only prerequisite is a duplicate role name, the severity is considered high. Patches This vulnerability has been patched in version 0.1.1 of the `auth-aws` plugin. Users are advised to upgrade to version 0.1.1 or later to remediate this vulnerability. Workarounds For users who are unable to upgrade to version 0.1.1 immediately, the most effective workaround is to guarantee that IAM role names are unique across all AWS accounts that could potentially interact with your OpenBao environment. This is the most critical mitigation step. Primary Mitigation: Audit your AWS organizations to identify and rename any duplicate IAM role names. Enforce a naming convention that includes account-specific identifiers to prevent future collisions. While removing wildcards from your `bound_iam_principal_arn` configuration is still recommended as a security best practice, it will not mitigate this vulnerability if duplicate role names exist. Credits This vulnerability was discovered and reported by Pavlos Karakalidis