5 juin 2026
It is Anthropic’s advanced cybersecurity AI model, part of the company’s expanded Project Glasswing rollout.
Up to nine Indian organisations across cyber, telecom, banking and finance have been onboarded so far for Project Glasswing, with major IT services firms excluded. India’s national cybersecurity agency, the Computer Emergency Response Team – India (CERT‑In) is among the public agencies expected to receive previews.
The government is actively negotiating to widen access.
Anthropic recently expanded Mythos access to 150 organisations across 15+ countries, emphasising entities that defend critical infrastructure such as power, water, healthcare, communications and financial systems. Existing global partners have already used Mythos to uncover over 10,000 high‑severity vulnerabilities in major codebases.
The controlled rollout reflects concerns about the model’s offensive‑security capabilities and the need to prevent misuse. India’s inclusion came after months of government‑level engagement with the US to ensure its critical infrastructure is not disadvantaged as other nations adopt Mythos‑class tools.
Check out my books, The Black Flames, The Flames Turn White, and crawlers, a poetry collection on KINDLE.