Mon, 2 Nov 2026
Anthropic · Security

Anthropic previews its powerful Mythos model for security partners only

Mythos has found thousands of zero-day vulnerabilities. Only 40 organizations get access through Project Glasswing.

Osamah Ismail · Founder & Editor-in-Chief
· 1 min read · Updated:
Anthropic previews its powerful Mythos model for security partners only
Image: Anthropic

What's new

On April 7, Anthropic released a limited preview of its new model Mythos. It is not generally available.

Instead, it is part of Project Glasswing, a defensive cybersecurity initiative.

Who gets access

  • 12 partner organizations will deploy it, and 40 organizations in total have access.
  • Partners include Amazon, Apple, Broadcom, Cisco, CrowdStrike, the Linux Foundation, Microsoft and Palo Alto Networks.

Why it matters

Anthropic says Mythos found "thousands of zero-day vulnerabilities, many of them critical," and many had gone unnoticed for one to two decades.

Because the model finds bugs so effectively, Anthropic wants defenders to use it first, before attackers can.

Background

Details of the model had leaked earlier under the name "Capybara" through an unsecured data cache, which Anthropic blamed on "human error."

Sources: TechCrunch, Anthropic
Osamah Ismail

Founder and editor-in-chief of Kurd Developers. Phone and tech news on YouTube, Instagram and TikTok.