AI tools help attackers exploit old bugs in PHP, researchers warn

In short: Researchers say AI is making it faster and easier to turn old PHP bugs into real attacks.

What's going on

Security researchers are warning about a growing pattern, AI tools are helping people dig up security problems that have been sitting in software for years. The example highlighted this week involves PHP, a widely used programming language that helps power many websites.

A “vulnerability” is a mistake in code that can let someone break in, like a bad lock on a door. Some of the PHP issues being discussed are not new. What is changing is how quickly people can go from reading old reports to building working attacks, with help from AI systems that can summarize technical details and suggest steps to try.

According to WIRED’s Kernel Panic newsletter, researchers used AI tools to help uncover years-old weaknesses in PHP. After that, AI-assisted attacks began. The concern is not only that more bugs will be found, but that the time between finding a bug and seeing it abused in the wild may shrink.

What to watch

Organizations that run PHP-based sites may need to patch faster and take old warnings more seriously, even if a bug seems obscure. Watch for more cases where AI lowers the skill needed to exploit known issues, especially in popular “building block” software that many sites rely on.

Source: Wired