AI security tools have successfully identified critical vulnerabilities in widely used software like PHP.

AI-discovered flaws in PHP, the language behind roughly 70 per cent of websites, forced governments from France to Australia to take sites offline. Researchers at Calif and Wiz found the vulnerabilities using OpenAI's GPT-5.6 Sol and Gemini 4 Argon. VulnCheck said one exploitation attempt came from an IP address in China. https://www.scmp.com/news/world/article/3371694/websites-go-dark-worldwide-ai-exposes-critical-security-flaws #China #Tech #AI

Two critical PHP vulnerabilities, present since version 5 from 2004, were uncovered by AI models. OpenAI's GPT-5.6 Sol, working with security firm Calif, found one, while Google's Wiz team used its Atlas agent running on Gemini 4 Argon to find more. The NSA warned that unpatched sites should be taken offline. https://techcrunch.com/2026/10/21/php-armageddon-openai-googles-wiz-find-ancient-holes-panicking-the-internet/ #AIagent #AI #GenAI #AISecurity