In brief
An AI model discovered a long‑hidden flaw in PHP, the scripting language behind 70% of the internet, sparking worldwide panic and emergency patch efforts. The bug, flagged by AI to Cyber‑security firm Wiz as the most severe ever found, raises concerns about the security of everything from Wikipedia to e‑commerce sites. OpenAI described the discovery as one of the most impactful security issues in recent memory, and government agencies are scrambling to contain the risk. BBC tech correspondent Joe Tidy explains how AI can unearth threats hidden in legacy code, and what this means for the future of web security.
When an AI system combed through millions of lines of code, it flagged a forgotten flaw in PHP, the server‑side language that powers more than 70 % of the web. Hackers could now exploit the bug to hijack it, steal data, or crash sites at a scale spanning everything from Wikipedia to thousands of e‑commerce platforms.
The vulnerability – deemed the “most severe” by cyber‑security firm Wiz and called “one of the most‑impactful security issues in recent memory” by OpenAI – had lurked in PHP for years. PHP developers rushed to release patches, and governments and agencies worldwide have issued emergency advisories and urged immediate updates.
Joe Tidy, BBC’s technology correspondent, reports that the discovery illustrates a striking synergy between artificial intelligence and cyber‑security, showing how AI can unearth risks that even seasoned programmers can miss.
To explore the full details, watch the in‑depth briefing on our virtual newsroom, or read more on the cyber‑security threat here: OpenAI and PHP patch brief.
