Artificial intelligence tools discover severe security bugs in core web platform

Security researchers using advanced artificial intelligence systems have uncovered major software flaws in PHP, the coding foundation behind roughly seven in ten websites. According to reporting by The Next Web, companies including OpenAI, Wiz, and Calif identified two severe vulnerabilities that could allow hackers to take control of web servers without needing a password. The discovery forced developers to hurry out emergency software updates to prevent potential global cyberattacks.

The situation grew chaotic after details of the flaws leaked online before official fixes were fully ready, forcing several major online services to temporarily pull their systems offline. While tech giants frame the breakthrough as evidence that artificial intelligence can protect the web by spotting weaknesses before criminals do, auditing firms involved in the project offered a more tempered view. As security engineers at Calif noted in their write-up, human expertise and deep knowledge of legacy code remain far more critical than relying solely on automated prompts.

Because the affected code had lingered quietly in web software for over a decade, this event highlights how deeply rooted digital vulnerabilities can be. What remains unclear is whether automated scanning tools will primarily help defend critical systems, or if rogue actors will adapt the same techniques to find and exploit secret flaws even faster. If artificial intelligence makes finding ancient bugs routine, the tech industry may soon face an overwhelming flood of urgent security patches that small teams simply lack the resources to maintain.