Calif researchers used artificial intelligence tools to uncover a critical security flaw in Java, a widely used programming language, in its memory functions. According to CyberScoop, Calif assigned the vulnerability a score of 9.2 out of 10 on the CVSS scale, above the threshold of nine for critical severity.
According to CyberScoop, the flaw affected hand-written processor instructions that accelerate Base64 decoding on chips supporting AVX-512; systems using ARM64 were not affected. Calif found that an input of 64 “A” characters followed by an equal sign caused a length miscalculation, writing 64 bytes into a buffer that could hold only 50.
According to CyberScoop, Although the program rejects the malformed input after the write, the surrounding memory is already damaged, and the corruption alone usually causes a crash. Calif said attackers could use heap grooming to arrange nearby memory and turn the corruption into control of a server.
According to CyberScoop, the group demonstrated unauthenticated code execution in Keycloak, Jira, Confluence, Jenkins and PingFederate, and showed a working break-in against Keycloak. Calif said researcher Quang Luong found the bug in about an hour while auditing OpenJDK with OpenAI’s Codex, using less than $200 in computing tokens.
At publication, CyberScoop reported no exploitation in the wild and said the related CVEs had not been added to CISA’s Known Exploited Vulnerability list. Elbaz urged organizations to treat the issue as serious exposure and apply compensating controls capable of blocking exploitation while patches were deployed.
Computer scientist Toby Walsh told the BBC Java’s role in managing and protecting memory made a flaw in that function particularly serious and unexpected. The BBC connected the disclosure with recent global technology outages, but reported Secretary of State Marco Rubio said there was “zero evidence” they were related.
The BBC said the NSA reported its internal communications went offline Tuesday because of scheduled routine maintenance and did not specify a connection to the Java flaw. The BBC reported OpenAI President Greg Brockman announced earlier in November that the company was shifting 25% of its staff toward cybersecurity demands.
Walsh told the BBC the Java flaw had remained hidden for 12 years and said AI was revealing obscure vulnerabilities faster, including ones humans might not have found alone.
Reader comments (0)
Bring evidence, ask questions, or challenge our reporting. Every comment is checked before publication.
No approved comments yet
Be the first to ask a thoughtful question or contribute useful context. We won't invent reader conversations.
Join the discussion
Members-only commenting
Everyone can read comments. To write, verify your email and sign in to your free member account.
Sign in or join freeNo newsletter subscription is required. Your email address remains private.