Researchers have discovered a serious flaw in the Java programming language, which is widely used in critical software in businesses, government and other sectors.
The issue, called Java4Shell, can be used to execute malicious code in a system with full access privileges, according to start-up Calif, which discovered the issue using OpenAI’s new GPT-6 Cyber model that’s not yet available to the public.
Oracle, which owns Java, released a patch for the flaw on Tuesday.
Java, released in 1995, is the “beating heart of the modern economy”, Calif said in an advisory.
“Banking, aviation, telecommunications, nearly every critical sector runs on Java, including national security systems,” the company said.
Calif last month uncovered a critical flaw in PHP that became publicly known earlier than planned, prompting organisations around the world to disable web-facing systems as they fixed the issue.
The Java bug affects organisations’ internal systems, and is more difficult to exploit, Calif said, but its presence could also leave unpatched systems vulnerable for years to come.
The issue is triggered on certain hardware when Java’s base64 decoder attempts to decode 64 “A” characters followed by an equals sign, according to Calif:
Critical software
The firm said it had used the issue to break into Keycloak, an authentication and login system widely used in large organisations.
Security agencies urged organisations to apply the patch immediately.
The discovery comes after days of outages affecting multiple unrelated systems, including Microsoft’s online business products, Apple services, banking services in the US, US airline flight control systems and a classified communications network used by the NSA, which have raised concerns around the stability of critical software.
Appearing in television interviews on Tuesday afternoon, director of national intelligence Jay Clayton said the NSA outage was due to a software update that he said was related to the PHP vulnerability.