Security teams and malicious actors leverage artificial intelligence at unprecedented speed

Cybersecurity researchers at Wiz recently reported that advanced artificial intelligence enabled them to discover critical, decade-old security flaws within PHP, the programming language that underpins a vast portion of the web. The discovered bugs allow attackers to seize control of web servers or read private memory without providing any login credentials.

Because PHP drives popular website tools such as WordPress and Wikipedia, these vulnerabilities indirectly expose millions of everyday web users to potential data theft and site vandalism. The discovery highlights a major shift in digital security: safety mechanisms that once depended on the sheer difficulty of building complex exploits are breaking down now that automated tools can work out those steps in a matter of hours.

PHP developers have released an official software update to fix the underlying memory issues, while major cloud service providers have deployed temporary shields for web hosts that cannot upgrade right away. The researchers coordinated their findings with software maintainers to ensure patches were ready before detailing the security gaps publicly.

It remains unclear how rapidly smaller website administrators will adopt the official patch, leaving many servers vulnerable in the interim. This situation raises a broader issue for digital infrastructure: as artificial intelligence lowers the technical skill needed to break complex software, security teams may struggle to distribute fixes faster than malicious scripts can scan for unpatched targets.