A major security flaw found using an OpenAI model is causing internet chaos (www.businessinsider.com)

A significant security flaw discovered using OpenAI's AI models poses a threat to approximately 70% of the world's websites, particularly those utilizing PHP, a popular open-source programming language. Cybersecurity researchers identified two "critical" vulnerabilities in PHP's infrastructure, prompting the National Security Agency's cyber division to advise web providers to take immediate action by either taking their systems offline or updating their software. OpenAI's president, Greg Brockman, warned that these exploits could lead to severe repercussions, including theft of private data and website manipulation. The vulnerabilities were reportedly recognized before being exploited by hackers, but the urgency of the situation has spurred a global response. Prominent websites, including government and media platforms, temporarily went offline to address the flaws. The discovery was part of an initiative called "Patch the Planet," which aims to strengthen digital infrastructure. As early attacks have already surfaced, Cloudflare's CTO emphasized the need for continuous mitigation efforts. This incident underscores the critical role AI models can play in detecting and addressing cybersecurity risks, highlighting the intersection of AI and cybersecurity in safeguarding digital environments.