Researchers have disclosed a critical memory-corruption vulnerability in OpenJDK, dubbed Java4Shell, which was identified using AI coding tools. By Greg Otto.
Security researchers at Calif Research have disclosed a critical memory-corruption flaw in OpenJDK, tracked as CVE-2026-114938. The vulnerability, nicknamed Java4Shell, allows unauthenticated attackers to potentially gain full control over servers by corrupting memory. The flaw resides in hand-written processor instructions used to optimize Base64 decoding. Specifically, the decoder miscalculates buffer lengths when processing a specific input of 64 ‘A’ characters followed by an equal sign, leading to an out-of-bounds write.
This issue specifically affects systems supporting the AVX-512 instruction set, while ARM64 systems remain unaffected. While the corruption typically causes a program crash, researchers demonstrated that ‘heap grooming’ techniques can turn the error into remote code execution. The group successfully demonstrated unauthenticated execution against Keycloak and authenticated execution against Kafka and Elasticsearch. Notably, the discovery was heavily accelerated by AI; researcher Quang Luong identified the bug in roughly one hour using OpenAI’s Codex for less than $200 in tokens.
Oracle has released a patch as part of a November update. While no exploitation in the wild has been reported yet, the speed at which AI can generate exploit chains suggests a high risk of rapid weaponization. Engineers should prioritize patching OpenJDK environments and monitor for tailored exploits targeting specific hardware configurations. Interesting read!
[Read More]