Anthropic Mythos AI expands to India: Which entities gained access? What we know | Mint

Indian entities have secured access to Anthropic's Mythos AI model as part of Project Glasswing expansion. Single-digit number of organisations granted permission to detect software vulnerabilities.

India has entered an exclusive club of nations with access to Anthropic's Mythos artificial intelligence model through the expanded Project Glasswing initiative, but the identities of the Indian organisations granted this privileged access remain shrouded in secrecy, raising questions about what the cybersecurity programme reveals regarding India's strategic positioning in global AI development.

A single-digit number of Indian entities spanning both governmental and private sectors have secured positions within the expanded cohort, according to sources familiar with the arrangement. Despite multiple inquiries, neither Anthropic nor the participating Indian organisations have disclosed which institutions hold this access, prompting speculation about which critical infrastructure operators or technology firms received the coveted inclusion.

The programme, described as "a collaborative effort to secure the world's most important software," targets organisations where cyberattacks could reverberate across national boundaries, potentially compromising the safety and operations of over 100 million people. "And many of the new partners are vendors, companies or non-profits that maintain codebases that are relied upon by lots of other organisations around the world, including governments," Anthropic stated in a June 2 announcement.

India's inclusion within this ecosystem reflects the country's positioning as both a consumer of advanced cybersecurity technology and a potential provider of software solutions relied upon globally.

For India's broader AI landscape, the arrangement represents validation that domestic organisations possess the sophistication and responsibility levels demanded by frontier AI developers for handling powerful security tools. This carries downstream effects: organisations gaining access to Mythos will develop hands-on experience with advanced AI-powered vulnerability detection, building domestic expertise that could inform India's own AI security initiatives and potentially shape the country's regulatory approaches to artificial intelligence.

Since the programme's April announcement, the initial 50 partners identified more than 10,000 high-severity or critical-severity security flaws across their codebases, demonstrating the practical impact of deploying AI against real-world security challenges.

"We believe it would be good for the world to have the option to slow or temporarily pause frontier AI development to enable societal structures and alignment research to keep up with the advance of the technology," the company said in a recent blog post titled "When AI builds itself."

The message underscores a growing tension within the AI industry: whilst companies like Anthropic provide advanced capabilities to select partners through programmes like Glasswing, they simultaneously advocate for global restraint in pushing AI development forward. For India, the question becomes whether access to Mythos enhances the nation's ability to manage increasingly capable AI systems, or whether the country's participation in such programmes obligates it to engage in broader conversations about AI governance and development timelines.