Kilambi News
Wednesday, November 18, 2026

AI model finds critical flaw in Java platform

Technology Tuesday, November 17, 2026 - Updated Nov 17, 2026 17:54

A security firm said on Tuesday that it had found a critical vulnerability in Java, the software platform used by government agencies and about 90 percent of Fortune 500 companies, using OpenAI's GPT-6 Cyber model.

Why it matters: Java runs inside banks, hospitals, airlines and technology giants, so a critical flaw coordinated across five allied governments points to a broad patching effort now under way.

Data as of Reporting of November 17, 2026, via UOL in Portuguese reporting The New York Times; single-source chain, graded C accordingly. No official advisory from Oracle or a government agency had been read at publication time.

Security firm Calif said on Tuesday that it had discovered a critical vulnerability in Java with the help of OpenAI’s GPT-6 Cyber model, one of several serious flaws the firm says it has found with AI models in recent months. Researcher Bruce Dang said the flaw will affect practically all large companies, and that he and colleagues have spent recent days coordinating a response with agencies in the United States, Canada, Australia, New Zealand and the United Kingdom, alongside companies including Google and Microsoft. Java is used by public bodies and about 90 percent of Fortune 500 companies, including banks, hospitals, airlines and technology giants.

Sources

  1. IA ajuda empresa a descobrir falha crítica no Java - UOL, reporting The New York Times - 2026-11-17
  2. A.I. Helps Find Major Security Bug in Java, a Software Platform Vital to the Internet - The New York Times - 2026-11-17