Researchers have leveraged artificial intelligence to uncover two previously unknown security flaws in PHP, the programming language powering roughly 70% of all websites worldwide. The findings highlight how AI-driven tools can accelerate vulnerability detection in widely used software, raising concerns about the scale of unpatched risks across the internet. Experts warn that such weaknesses could expose millions of sites to exploitation if left unaddressed. The discovery underscores the growing role of machine learning in cybersecurity, even as it prompts questions about whether developers can keep pace with automated threat discovery.
AI was used to pinpoint two vulnerabilities in the PHP language, used by about 70 percent of websites.
