字号·· | 护眼
罗塞塔简报

AI发现Java严重内存漏洞 甲骨文发布安全补丁

点「原文对照」整页切到原文,或双击某段只看那段的原文。

网络安全初创公司Calif利用OpenAI尚未发布的GPT-6 Cyber模型,在Java编程语言中发现了一个名为Java4Shell的严重内存漏洞。该漏洞允许在Java的内存系统中远程执行非预期命令,据报已隐匿12年未被发现。Java所有者甲骨文公司尚未确认这一具体漏洞,但已于2026年11月17日发布了包含安全修复的关键补丁更新。目前尚不清楚是否有任何公司或平台受到影响。

AI Uncovers Critical Memory Flaw in Java as Oracle Releases Security Patch Cyber-security start-up Calif used OpenAI's unreleased GPT-6 Cyber model to discover Java4Shell, a critical memory vulnerability in the Java programming language. The flaw allows remote execution of unwanted commands in Java's memory system and reportedly went undetected for 12 years. Java owner Oracle has not acknowledged the specific bug, though it released a critical patch update containing security fixes on November 17, 2026. At this stage, it remains unclear whether any companies or platforms have been affected.

这一发现发生在苹果、微软、美国银行、美国航空网络和美国国家安全局近期发生故障之后,但美国国务卿马尔科·鲁比奥表示,没有任何证据将这些事件联系起来。专家警告称,未修补的漏洞可能威胁关键金融和基础设施系统。

The discovery follows recent outages at Apple, Microsoft, US banks, the US aviation network, and the US National Security Agency, though US Secretary of State Marco Rubio stated there is zero evidence linking these incidents. Experts warn that unpatched flaws could threaten critical financial and infrastructure systems.

阅读媒体原文 ↗