Java4Shell: Critical OpenJDK Flaw Enables Remote Code Execution in Jenkins, Jira, Confluence and Keycloak

A critical vulnerability in OpenJDK, the open-source implementation of Java, dubbed Java4Shell, can let unauthenticated attackers remotely execute code on popular enterprise applications, including Jenkins, Atlassian Jira, Confluence, and Keycloak.
Java is one of the most widely used programming platforms in the world, powering enterprise applications, web services, financial systems, and critical business infrastructure. Because many applications share the same Java runtime, flaws in its core components can affect many different and unrelated applications.
The vulnerability
Tracked as CVE-2026-114938, the flaw is a memory-corruption bug in Java’s hardware-accelerated Base64 decoding. It can be exploited to leak sensitive information from memory or to achieve remote code execution (RCE).
| Rating source | Score | Severity |
|---|---|---|
| Calif (CVSS 4.0) | 9.2 | Critical |
| Oracle (CVSS 3.1) | 8.1 | High |
How it works
Java’s Base64 decoder processes encoded data commonly found in authentication, file uploads, and network communications. According to Calif, the security firm that disclosed the flaw on Tuesday, specially crafted Base64 data can make the decoder overwrite parts of Java’s memory, potentially letting attackers steal sensitive information or run malicious code.
The researchers say the memory corruption can be triggered with just 64 “A” characters followed by a single equals sign (=). Older OpenJDK 17 versions need 128 “A” characters followed by the equals sign.
Exploitation conditions
- The flaw only affects a hardware-accelerated decoding function used on certain x86-64 processors with AVX-512 VBMI support.
- There is no universal exploit. Attackers must find an application-specific way to get malicious Base64 data to the vulnerable decoder.
Affected Java versions
Java4Shell affects OpenJDK 17.0.6 and later, and Java distributions built from the affected code, including:
- Oracle Java SE
- Amazon Corretto
- Eclipse Temurin
- Microsoft Build of OpenJDK
- Azul Zulu
- Red Hat OpenJDK
Applications exploited by researchers
Calif exploited the flaw to achieve remote code execution against widely used Java applications running with their default configurations:
| Type | Applications |
|---|---|
| Unauthenticated RCE | Keycloak, Atlassian Jira, Atlassian Confluence, JFrog Artifactory, Jenkins, PingFederate |
| Authenticated RCE | Apache Kafka Broker, Apache Solr, Elasticsearch, Gerrit, Spinnaker |
Calif also exploited the flaw in sample applications built with the Apache Struts and Spring Boot frameworks.
Found with AI in about an hour
According to Calif, researcher Quang Luong found the flaw using OpenAI’s Codex AI coding assistant during an audit of OpenJDK’s internals.
- He found the flaw in about one hour.
- He built a first working exploit in another two hours.
- The total AI usage cost was less than $200.
Calif says AI agents also helped develop additional application-specific exploit chains.
Proof of concept
Calif has published a proof of concept showing the memory corruption, but has not released the weaponized exploits used against the affected enterprise applications.
Patch status
Oracle released patches for CVE-2026-114938 on November 17 as part of its November Critical Security Patch Update, which fixes 655 security vulnerabilities across its product portfolio. Oracle recommends applying the available patches as soon as possible.
Temporary mitigations
For anyone who can’t update their Java runtime right away, Calif recommends:
- Disabling the vulnerable hardware-accelerated Base64 function by starting OpenJDK with
-XX:-UseBASE64Intrinsics - Or, on x86-64 systems, disabling AVX-512 instructions with
-XX:UseAVX=2
These workarounds are temporary. All developers should update their Java runtimes to the latest version.



