Most of us never think about PHP, but it’s working quietly behind a huge number of the websites we visit every day. This week, experts found two dangerous weak spots hiding inside it, and the way they found them has everyone talking: they used AI.
A security company called Calif spotted the first problem back in August, with help from an AI made by OpenAI, the company behind ChatGPT. A second team at Wiz, a company owned by Google, found the problems on its own using AI tools too. Instead of shouting about it, both teams quietly helped PHP build a fix. On Wednesday, PHP released emergency updates.
The problem is huge. Every version of PHP from the last ten years has at least one of these weak spots, and Wiz says PHP runs about seven out of every ten websites. A hacker could have used them to take over a site without ever needing a password.
That’s why websites are going dark. The NSA, a U.S. security agency, told site owners that if they can’t fix their site right away, they should turn it off until they can. Wikipedia switched off some of its systems on Thursday just to be safe, and more than 200 groups in Japan shut down for repairs.
So is AI the hero here, or the villain? OpenAI says AI found these holes before the bad guys did. But experts warn that the same AI could also make it easier for hackers to break in. AI helped stop this problem, and AI is also what makes it so scary. That’s the question people will keep arguing about long after the fixes are done.
If you run a website, update PHP today.
Sources:
- TechCrunch: PHP Armageddon: OpenAI, Google’s Wiz find ancient holes, panicking the internet
- CNN: AI finds cybersecurity issue affecting 70% of the internet, forcing websites to shut down
- The Next Web: PHP rushes out emergency patches for two critical bugs AI found first
- Quartz: PHP releases patches for two critical flaws that AI helped researchers find
- PPC Land: OpenAI says GPT-5.6 Sol found a PHP flaw before attackers did
- AI Buzz News for October 22, 2026
Leave a Reply