Red Hat AMQ Broker 7's Artemis messaging subsystem uses ObjectInputStreamWithClassLoader for deserialization, which supports allow-list and block-list filtering. However, by default, both lists are empty, causing the isTrustedType() method to return true for all classes, effectively allowing deserialization of any class. This behavior can be exploited to deserialize untrusted data, potentially enabling attackers to execute arbitrary code or cause denial of service. The vulnerability is identified as CVE-2026-86404 with a CVSS 3.1 score of 8.8 (high).

The vulnerability allows deserialization of arbitrary classes, which can lead to unauthorized code execution, modification of application data, and denial of service through resource exhaustion. The impact affects confidentiality, integrity, and availability of the affected system. Attackers with low privileges and no user interaction required can exploit this over the network.

A fix is not explicitly stated as available yet. Patch status is not yet confirmed — check the Red Hat advisory at https://access.redhat.com/security/cve/CVE-2026-86404 for current remediation guidance. Meanwhile, Red Hat recommends configuring a restrictive deserialization allow-list on every pooled-connection-factory resource to permit only expected classes. For example, if message-driven beans expect objects of type com.yourapp.OrderEvent, set the allow-list to "com.yourapp" to deny deserialization of all other classes.