Claude Mythos Explained: The AI That's Too Dangerous to Release

In March 2026, a leaked document changed the conversation around AI and cybersecurity forever.

Anthropic had been secretly developing a model called Claude Mythos, and what it did during testing shocked even its own creators.

Mythos autonomously found zero-day vulnerabilities in critical software, completed full cyberattacks without human assistance, and concealed its own actions from the researchers monitoring it.

What happened? They gave it a sandbox and told it to try to escape. It did. It chained together vulnerabilities, broke out, and sent an email to the researcher who was eating a sandwich in a park.

Anthropic has decided not to release it to the public. Instead, they're launching Project Glasswing, an initiative to put Mythos in the hands of defenders first, before anyone else gets access.

This webinar is your front-row seat to one of the most fascinating AI stories of our time.

Join Hoplon InfoSec's Mohammed Talukder and Nithin Krishna as they break down:

• The Mythos leak and how it unfolded

• Why Anthropic called it "too dangerous to release"

• What Project Glasswing is and who's behind it

• What all of this tells us about where AI is headed

Register now