What is Mythos and why has it been blocked

▤ Summary

Anthropic has blocked access to its most advanced Mythos models after U.S. government restrictions raised concerns about misuse by foreign actors. The article explains that Mythos can identify software vulnerabilities and generate exploits with limited human input, which makes it useful for both defenders and attackers. It also notes that Anthropic initially shared the model with a limited set of partner organizations for defensive security work. The story highlights a wider AI security race, where vendors are building tools to find zero-day flaws faster while cybersecurity teams try to keep up.

▦ Linked Companies

Anthropic

$10M to $25M