Claude Mythos: Anthropic's Leaked Next-Gen Model That Has Governments Worried About Cybersecurity

On March 26, 2026, a CMS configuration error exposed nearly 3,000 unpublished Anthropic assets — including a draft blog post describing Claude Mythos, a new model tier above Opus that Anthropic calls 'by far the most powerful AI model we have ever developed.' The leaked documents reveal a 'Capybara' product tier with dramatically higher scores in coding, reasoning, and cybersecurity, alongside private government warnings that Mythos makes large-scale cyberattacks 'significantly more likely in 2026.' Cybersecurity stocks dropped on the news.