tech
April 7, 2026
Anthropic debuts preview of powerful new AI model Mythos in new cybersecurity initiative
The new model will be used by a small number of high-profile companies to engage in defensive cybersecurity work.
TL;DR
- Anthropic has released a preview of its new frontier model, Mythos.
- Mythos will be used by partner organizations for defensive cybersecurity work and to secure critical software.
- The model has identified thousands of zero-day vulnerabilities, many of them critical.
- Project Glasswing involves over 40 partner organizations, including major tech companies.
- Mythos is a general-purpose model with strong coding and reasoning skills.
- The preview is not generally available and partners will share their findings.
- Discussions are ongoing with federal officials regarding Mythos's use.
- News of Mythos (previously codenamed 'Capybara') was leaked last month.
- The leaked document described 'Capybara' as larger and more intelligent than previous Opus models.
- Anthropic previously experienced a data security incident involving leaked source code.
Continue reading the original article