Anthropic's Claude Mythos Finds Thousands of Zero-Day Flaws Across Major Systems
Briefly

Anthropic's Claude Mythos Finds Thousands of Zero-Day Flaws Across Major Systems
"Mythos Preview has already discovered thousands of high-severity zero-day vulnerabilities in every major operating system and web browser, including a now-patched 27-year-old bug in OpenBSD."
"In one instance, Mythos Preview autonomously created a web browser exploit that chained together four vulnerabilities to escape the renderer and operating system sandboxes."
"The model solved a corporate network attack simulation that would have taken a human expert more than 10 hours, showcasing its advanced capabilities."
"Mythos Preview managed to follow instructions from a researcher to escape a secured 'sandbox' computer, indicating a 'potentially dangerous capability' to bypass its own safeguards."
Anthropic launched Project Glasswing, utilizing its Claude Mythos model to enhance cybersecurity by identifying vulnerabilities. A select group of organizations, including major tech companies, will use this model to secure software. The initiative arose from the model's advanced coding capabilities, which can exceed skilled human performance in vulnerability detection. Mythos Preview has already uncovered numerous high-severity zero-day vulnerabilities across major systems. Notably, it demonstrated the ability to autonomously exploit vulnerabilities and escape sandbox environments, raising concerns about its potential misuse.
Read at The Hacker News
Unable to calculate read time
[
|
]