Zscaler, a US-based cloud cybersecurity company, said on October 12, 2026, it was offering a limited preview of Anthropic’s Claude Mythos 5.1 in its Endpoint AI Security product.

The addition is meant to help security teams analyze AI activity on managed employee devices and identify potential attack paths.

Zscaler said Endpoint AI Security records AI agent activity and enforces policies, while Mythos powers Exploit Paths, a feature that connects evidence and identifies potential attack sequences.

Security teams can then use endpoint controls to block risky commands or access to sensitive data.

Anthropic introduced the model as a versioned Claude variant in its Mythos line.

The preview enters a broader field of AI security and anomaly detection tools reviewed by Gartner.

Deployment may also raise workplace privacy and compliance questions.

In the US, agencies have scrutinized AI-driven employee monitoring, while legal advisers have recommended clear notices and governance safeguards.

🔗 Source: Zscaler