Calif uses AI to find critical Java4Shell vulnerability in Java
Covered by 2 sources

Cybersecurity startup Calif employed OpenAI's GPT-6 Cyber model to uncover a decade-old Java flaw called Java4Shell that allows remote code execution. The vulnerability impacts many Fortune 500 companies and vital systems. Oracle has issued a patch, but unpatched systems remain at risk.
What happened
Cybersecurity startup Calif used OpenAI's GPT-6 Cyber model to identify a serious security flaw in Java's base64 decoder. The vulnerability, known as Java4Shell, has existed for over a decade and allows remote command execution and system crashes. Oracle has released a patch to address the issue. Many organizations using unpatched Java systems remain at risk.
Key details
- Java is used by 90% of Fortune 500 companies.
- The flaw affects critical infrastructure, including Keycloak login systems.
- The vulnerability allows attackers to exploit memory systems for unauthorized access.
- Oracle issued a critical patch update addressing the flaw.
- The GPT-6 Cyber model used for the discovery is not publicly available.
Why it matters
Organizations that rely on Java, including major corporations and critical infrastructure providers, face a security risk if they do not apply the patch. Unpatched systems are vulnerable to remote attacks that could disrupt operations or compromise sensitive data. IT administrators and security teams must act quickly to update their systems to prevent exploitation.

