5 Hrs
Source:
منصة تك عربي
Anthropic introduces Claude Mythos for vulnerability detection
Anthropic has revealed that its advanced AI model, Claude Mythos, managed to identify over 10,000 severe security vulnerabilities within just one month during a pilot project called "Glasswing." The model assisted its partners in discovering hundreds of flaws, including around 2,000 vulnerabilities for Cloudflare and 271 for Mozilla, achieving up to ten times the results of traditional tools. Despite its impressive capabilities, the company notes that it restricted the model's usage out of concern that it could be misused for cyberattacks, amid fierce competition among AI companies striving for supremacy in cybersecurity.
Size: 120%Text Size
Notice: This Is an AI-Generated Summary
Comments (0)