AI identifies critical security flaws in PHP, a language used by millions of websites

Security experts have identified two critical vulnerabilities in PHP, a widely used open-source programming language that powers nearly 70% of websites. The flaws were discovered using an OpenAI model, demonstrating the potential of artificial intelligence to uncover security issues that have gone unnoticed for decades. These vulnerabilities could allow hackers to take control of web servers without needing passwords or other credentials. While AI-generated reports have often been met with skepticism due to their tendency to highlight minor bugs or false alarms, this discovery underscores the growing role of AI in cybersecurity. The findings highlight the fragility of the modern internet, which relies heavily on open-source software. Cybersecurity professionals are now on high alert, anticipating a potential wave of attacks exploiting these vulnerabilities in the coming weeks. The discovery has sparked discussions about the reliability of AI in identifying critical security risks. Some experts remain cautious, noting that while AI can uncover significant issues, it is not infallible. Nonetheless, the incident marks a pivotal moment in the intersection of AI and cybersecurity, with implications for the future of online security. As the tech community scrambles to address these flaws, the incident serves as a stark reminder of the vulnerabilities inherent in the digital infrastructure that underpins the modern world.