Apple Alerted to macOS Security Vulnerability Uncovered With AI Tool

Anthropic precocious announced Project Glasswing, an inaugural that enables tech companies similar Apple to usage its caller frontier AI exemplary Claude Mythos Preview to find information vulnerabilities crossed operating systems and web browsers.


The Wall Street Journal contiguous reported that researchers astatine cybersecurity steadfast Calif utilized Claude Mythos Preview to uncover a caller macOS information vulnerability past month. Specifically, they utilized the exemplary to constitute codification that links unneurotic 2 macOS bugs successful a mode that resulted successful what is known arsenic a privilege escalation exploit.

The information researchers said the exploit would not person been imaginable with Mythos alone, arsenic it inactive required their quality expertise connected top, but it nevertheless proves that AI tin assistance with discovering bundle vulnerabilities.

Apple said it was reviewing Calif's study to validate the findings.

"Security is our apical priority, and we instrumentality reports of imaginable vulnerabilities precise seriously," an Apple spokesperson told The Wall Street Journal.

It is unclear if Apple has already patched the exploit. Apple's information notes for the macOS 26.5 update released this week notation a hole for a kernel-level vulnerability, and it credits Calif and Anthropic for discovering it. Yet, the study said that Calif lone met with Apple this week and suggested that a hole was inactive coming.

We person reached retired to Apple for comment.
This article, "Apple Alerted to macOS Security Vulnerability Uncovered With AI Tool" archetypal appeared connected MacRumors.com

Discuss this article successful our forums