
AI uncovers critical PHP security flaw with remote code execution risk
The story in brief
Researchers at Calif, with OpenAI, found the first PHP2Shell flaw on August 11, using AI to uncover vulnerabilities. The flaws, which allow remote code execution without authentication, were deemed among the most severe in PHP history. Wiz and the PHP Foundation warned of growing AI-driven security risks, prompting global patches and shutdowns.