AI Development

CERT-In secures access to Anthropic Claude Mythos for vulnerability testing

November 4, 2026 · Medium importance · deployment

India’s Computer Emergency Response Team (CERT-In) obtained access to Anthropic’s Claude Mythos model and began controlled experiments on software vulnerability identification. The model had previously demonstrated advanced capabilities in discovering zero-day flaws. Access occurs under Anthropic’s restricted Project Glasswing framework for vetted cybersecurity partners. This represents expanded official deployment of a high-capability cyber-focused model within an Indian government agency.

Date
November 4, 2026
Importance
Medium 3/5
Category
deployment
Axis of Change
Deployment Expansion
Organizations
Anthropic CERT-In
Models Affected
Claude Mythos
  1. 01 indianexpress.com