@igus_ai: Apple spent 5 years building a protection to shield its operating system. Three researchers bypassed it in 6 days using…
Summary
Three security researchers bypassed Apple's MIE protection in six days using AI model Claude Mythos, demonstrating a rare exploit that requires no traditional malware and gains root access on macOS. The full paper will be published after Apple releases a patch.
View Cached Full Text
Cached at: 05/16/26, 09:22 PM
Apple spent 5 years building a protection to shield its operating system.
Three researchers bypassed it in 6 days using Claude Mythos.
Apple designed MIE as the major security defense for the M5 and A19 chips, a layer created specifically to block most modern hacks even if there was a bug inside the system.
According to the company itself, it prevented all publicly known exploits in modern iOS.
But the Calif team found another route.
They didn’t break the protection directly. They went around it.
And the craziest part is the speed:
→ they discovered the bug on April 25 → by May 1 they already had a functional exploit developed with the help of Claude Mythos.
An extremely rare attack that doesn’t need to modify critical memory or execute malware in the traditional way.
Just normal system calls from an unprivileged account, until gaining root access in macOS.
And they personally went to Apple Park to deliver the technical report.
The full paper has 55 pages and will be published when Apple releases the patch.
Possibly the cybersecurity story of the year.
Polymarket (@Polymarket): JUST IN: Security researchers claim to have breached macOS using Claude Mythos.
Similar Articles
@cryptopunk7213: claude mythos just broke Apple's $2 billion defense system. it did so by discovering a completely different attack vect…
Claude Mythos AI discovered a novel attack vector that bypassed Apple's M5 chip defense system in five days at a cost of $35K, producing a 55-page report delivered to Apple. The exploit poisons data ingested by the chip, evading Apple's MIE system.
Pardon MIE?
A three-person team with AI assistance bypassed Apple's Memory Integrity Enforcement (MIE) in five days using two bugs and a clever idea, demonstrating a significant vulnerability in Apple's hardware-based kernel memory safety stack.
The first public macOS kernel memory corruption exploit on Apple M5 was built with Mythos Preview's help, and it only took 5 days.
Calif researchers, with help from the AI tool Mythos Preview, built the first public macOS kernel memory corruption exploit on Apple M5 hardware, bypassing MIE. The exploit chain took 5 days and will be fully disclosed after Apple fixes the vulnerabilities.
Anthropic's Mythos Just Helped Find macOS vulnerability That Could Break Apple's Security Protections
Anthropic's Mythos AI model helped cybersecurity firm Calif discover two previously undocumented macOS vulnerabilities that could bypass Apple's memory integrity enforcement, demonstrating the model's offensive capabilities under controlled access via Project Glasswing.
First public macOS kernel memory corruption exploit on Apple M5
Researchers from Calif disclosed the first public macOS kernel memory corruption exploit on Apple M5, bypassing the MIE hardware mitigation and achieving local privilege escalation on the latest macOS version.