OpenAI GPT-5.6 Sol aided researchers finding 2 PHP security flaws
Researchers using OpenAI's GPT-5.6 Sol uncovered two big security flaws in PHP, a programming language that powers nearly 70% of all websites.
The bugs, which have been around since 2004, were flagged through OpenAI's Patch the Planet program and reported to the PHP Foundation by cybersecurity firm Calif.
NSA advises keeping vulnerable sites offline
The flaws are so serious that experts called it one of the worst ever found. It could let hackers take control of affected servers and steal data or mess with services.
Big names like Wikipedia and government sites even went offline for a bit to patch things up.
The NSA recommends keeping vulnerable sites down until they're fully updated, and this whole episode is a reminder that while AI can help spot hidden problems, we need to use it carefully and responsibly.