AI Finds Old PHP Bug
Millions of attack attempts targeted vulnerable servers within days of the disclosure.
By Fatima Davis ·
Websites worldwide are facing a surge of automated attacks following a new security disclosure.
Researchers from Wiz Research and OpenAI revealed critical flaws in PHP on Wednesday. PHP is a programming language that powers approximately 70 percent of global websites. The teams stated that the bugs have existed in the code for decades but were difficult to exploit manually. They noted that AI models made the complex chain of steps required for exploitation practical for attackers.
OpenAI labeled the issue one of the most impactful vulnerabilities in recent memory. Wiz Research warned that successful attacks could give hackers control over servers, allowing them to steal data or deface sites. The firms developed a patch before making the findings public so organizations could apply a fix immediately.
Cloudflare head Matthew Prince reported that his company detected tens of thousands of exploit attempts in the first 45 minutes after the announcement. By Thursday, that number had risen to the millions. This rapid escalation has prompted a global response from government and private sector leaders.
The U.S. National Security Agency advised users to maintain heightened awareness. The U.K. National Cyber Security Center urged organizations to take immediate action and update to the latest patched version. China's Industry Ministry called for a review of security gaps. In Japan, Prime Minister Sanae Takaichi issued an urgent request for action and suggested that temporarily suspending websites could reduce risk.
Several services have already gone offline. Dozens of British and Finnish government services were taken down this week. Many local government sites in France were briefly pulled from the internet. The quick removal of these sites indicates that operators are prioritizing defense over availability while the patch is applied.
This event marks a shift in how AI is viewed in cybersecurity. Previously, reports of autonomous agents cheating in evaluations caused concern. Now, AI technology is credited with helping identify and mitigate a widespread threat that human researchers might have missed.
Source: NPR
