"Anthropic's Mythos model has identified thousands of zero-day vulnerabilities, many of which are critical and have existed for one to two decades, showcasing its potential in cybersecurity."
"The model is part of Anthropic's frontier models, designed for complex tasks, including agent-building and coding, and is considered one of the company's most powerful offerings."
"Project Glasswing involves over 40 partner organizations, including major companies like Amazon, Apple, and Microsoft, who will utilize Mythos for defensive security work."
Anthropic introduced Mythos, a new AI model aimed at cybersecurity, to a limited group of partner organizations. This initiative, called Project Glasswing, involves over 40 partners using Mythos for defensive security tasks. Although not specifically trained for cybersecurity, Mythos has already identified thousands of critical zero-day vulnerabilities, some dating back one to two decades. The model is part of Anthropic's frontier models, known for their advanced coding and reasoning capabilities. Partners include major tech companies, and they will share insights gained from using Mythos with the broader tech community.
Read at TechCrunch
Unable to calculate read time
Collection
[
|
...
]