AI-discovered vulnerability puts millions of websites at risk of getting hacked

Millions of websites around the world could be at risk of getting hacked. This after one of the most severe online bugs ever created was discovered with the help of AI. All right, Fox 5's Stephanie Bertini live in the newsroom with the latest on the race to patch this flaw. Stephanie. Steve, Natasha, new concerns tonight about web security after AI helped discover major risks with websites around the world. PHP, a widely used general purpose scripting language geared towards web development, became a major concern after a warning from the FBI's cyber division and the organization behind the language. PHP powers the vast majority of websites around the world, and AI technology found flaws. As a result, many organizations worldwide had to address the problem immediately. Security advisories yesterday urged PHP users to patch two critical vulnerabilities or shut down their sites if they couldn't update them immediately. That's because it was determined that there was potential for bad actors or hackers to access sites running PHP. Now, all of it highlights the power of AI when it comes to Web security. And experts say the big question now is if AI could be used in this case for good, how can it be used for bad? We talked to an expert about the concerns. The interesting part is that we're constantly, I think, in many ways, in the last weeks and months, talking about AI as a negative. In this situation, AI is a positive. This is something where AI as a software is being used to review for potential vulnerabilities, and one was, in fact, identified. So this is showing both sides of the coin. I think the question more widely being asked is if this is software that could find that, And it would be at the moment used for good. What happens when AI agents of this quality and level, which are not necessarily available to the open public, but if it was, how could it be used in order to create a negative situation? Yeah, a lot of questions. PHP powers about 70% of the world's websites. And so it was an immense possible security exposure. Patches are available now. But the big takeaway here is that if AI found the problem and not humans, what could be next?