The key facts
- Startup Calif used OpenAI's GPT-6 Cyber model to discover a vulnerability in Java
- Java is used by roughly 90% of Fortune 500 companies, including banking, healthcare, aviation, and government agencies
- Researchers note that AI finds vulnerabilities faster than specialists can verify them
AI finds a vulnerability in one of the most widely used programming languages
Cybersecurity startup Calif reported that it discovered a serious vulnerability in Java using OpenAI's recently released GPT-6 Cyber model. The New York Times reported this, and the Anadolu agency cites its article. Java underlies a significant part of the global internet: the language is used by roughly 90% of Fortune 500 companies and plays a notable role in banking, healthcare, aviation, and technology. Government agencies also use it, including structures responsible for national security.
What is known about the finding
Calif senior researcher Bruce Dang, in a comment to The New York Times, called Java 'the beating heart of the modern economy' and added that 'virtually all major companies' will be affected by this flaw. Calif had previously used leading AI models to uncover a series of serious software vulnerabilities; the latest finding concerns one of the most widely deployed programming languages in the world.
Broader context
According to an assessment by The New York Times cited by Anadolu, this finding is further evidence that increasingly capable AI systems are transforming cybersecurity research: they identify weaknesses in code at a pace that may exceed specialists' capacity to verify and fix the problems. The discovery underscores the growing concern among security experts and technology companies about the consequences of AI-assisted vulnerability research. The details of the vulnerability itself — its technical nature, remediation status, and patch timeline — are not disclosed in the publication.

