CVE-2026-89238

Apache WSS4J before 1.6.17 and 2.0.x before 2.0.2 improperly leaks information about decryption failures, making it easier for remote attackers to recover the plaintext form of a symmetric key.

An authentication bypass in the DOM security processor in Apache WSS4J allows unauthenticated remote attackers to forge authenticated SOAP messages via a crafted unsigned SAML sender-vouches assertion containing an attacker-controlled key.

In the StAX streaming WS-SecurityPolicy validator, certain relative or unsupported XPath expressions can be converted into paths that never match the actual XML element path. A remote SOAP peer may therefore send a required element without the expected signature or encryption.

Apache WSS4J stores the Nonce of each accepted UsernameToken as raw base64 text, but authentication decodes that text and uses the bytes. The same bytes can be written as base64 in several ways. An attacker who captured an authenticated request could re-send it with a space added to the Nonce: the password digest still verified, but the token no longer matched the remembered one, so the replay was accepted. Since a UsernameToken does not cover the message body, the captured token could then be reused on requests of the attacker's choosing until it expired. Affects deployments with a nonce replay cache configured, as Apache CXF has by default, and only tokens using a password digest. The cache is now keyed on the decoded Nonce. Users are recommended to upgrade to versions 4.0.2 or 3.0.6 or 2.4.4, which fix this issue.

Apache WSS4J accepted attacker-controlled derived-key lengths and offsets without adequate bounds. This could permit cryptographically weak keys or excessive CPU and memory consumption when processing crafted WS-Security messages.