Skip to content
澎湃新闻 · 检索「AI风险」· 澎湃新闻记者 李怡彬·Indexed by this site · Originally published· Event →· Its place in the scan record →

Large-scale system failures occur in succession in the United States; experts say AI-accelerated vulnerability discovery forcing emergency fixes may be the main cause

释新闻| AI 发现漏洞的速度,正在超过美国修复软件系统的能力?

AI summary

Within one week, the United States saw a series of large-scale incidents, including a Microsoft service outage, a nationwide suspension of flight departures, and a National Security Agency (NSA) system failure. Officials said none were cyberattacks, and experts said the main cause may be that AI rapidly discovered a large number of vulnerabilities, forcing emergency fixes. Director of National Intelligence Clayton said the NSA failure was related to how it handled a PHP vulnerability; the PHP2Shell vulnerability was discovered by researchers last month using AI tools, and after the patch was installed, some systems could not be restored. The classified communications network NSANet had been restored as of the 18th. In October alone this year, Microsoft fixed 1,407 security vulnerabilities, the most in the company's history, compared with about 170 a year earlier. Experts also noted that emergency patches are often not fully tested and can change core components such as the kernel, and may themselves become a source of failure; on November 5, the EU cybersecurity emergency response team warned that the AI-automated penetration testing tool Artex has been used by some threat actors for cyberattacks.

Read originalthepaper.cn