Security researchers recently turned to artificial intelligence to hunt down long‑standing security flaws in PHP, the popular web scripting language used by millions of websites worldwide. By feeding AI models with code samples and known vulnerability patterns, the researchers were able to identify several bugs that had persisted in the language for years.
Once the vulnerabilities were exposed, the same AI tools were repurposed to automate attacks against PHP‑based applications. The automated scripts could quickly generate exploit payloads, test for successful code execution, and propagate through vulnerable sites with minimal human intervention.
PHP remains a cornerstone of the web, powering a large share of the internet’s content management systems and e‑commerce platforms. The discovery of these dormant bugs underscores the importance of continuous code review and timely patching, especially when new technologies can accelerate both detection and exploitation.
While the research team did not disclose specific details about the affected PHP versions or the exact nature of the bugs, the incident highlights a growing trend: AI is becoming a double‑edged sword in cybersecurity, offering powerful tools for both defenders and attackers.
Industry experts advise that developers maintain up‑to‑date PHP releases, apply security patches promptly, and employ automated scanning tools to detect potential weaknesses before they can be weaponised by malicious actors.
<small>Source: Wired — read the original story there.</small>
