Anthropic is expanding its Claude Mythos cybersecurity AI programme to more than 15 countries, including India, granting select organisations preview access under its Project Glasswing initiative focused on strengthening cyber defence and protecting critical infrastructure.
The company said it is adding about 150 organisations, taking total participation to around 200 across 15+ countries, marking its first major expansion beyond the US and UK. The rollout spans financial services, cybersecurity, tech, healthcare, communications, energy, water and hardware sectors.
Participants include Okta, Samsung, SK Hynix, SK Telecom, Euroclear, Intercontinental Exchange (NYSE owner) and Swift, along with NATO and the EU Agency for Cybersecurity (ENISA), as well as Microsoft, Apple, Oracle, JPMorgan Chase and CrowdStrike.
Mythos, launched in April as a preview for about 50 mostly US organisations, is not publicly available due to its advanced vulnerability-finding capabilities. It is designed to detect software flaws and strengthen code security.
Anthropic says participants have already identified over 10,000 high or critical vulnerabilities under Project Glasswing.
India is included among new markets, though participants remain undisclosed; earlier reports suggested TCS, Infosys and CERT-In. Access requires strict eligibility checks.
Anthropic opened its Bengaluru office in February, calling India its second-largest Claude market, with rising enterprise adoption; firms like Air India, Razorpay and Cognizant use Claude, and Siddiq Zaman leads India partnerships.
Separately, Anthropic has confidentially filed for a US IPO, with reports suggesting a potential near $1 trillion valuation amid strong AI investor demand.
