Anthropic expands Claude Mythos access to India as enterprise and partner ecosystem grows

The cybersecurity-focused AI model is now available to organisations across 15 additional countries, including India.

Anthropic has expanded access to Claude Mythos, its cybersecurity-focused AI model, to organisations across 15 additional countries as part of Project Glasswing, the company's initiative focused on identifying software vulnerabilities and security weaknesses before they can be exploited by attackers.

While Anthropic did not disclose the newly added countries, a report by the Financial Times said the latest expansion includes India among the 15 markets granted access to the programme.

Earlier this year, Anthropic provided preview access to a group of technology companies, financial institutions and government agencies in the United States.

According to the company, participating organisations used the model to analyse software code and identify security flaws, with over 10,000 critical vulnerabilities detected during the initial phase.

The latest expansion brings the programme to over 150 organisations globally.

The company said participating organisations span sectors including power, water, healthcare, communications and hardware, where software vulnerabilities can have significant operational and national security implications.

It warned in its blog, “What each partner has in common is that a successful attack on their codebase could be catastrophic. For most partners, we estimate that a major attack could affect more than 100 million people, with important ramifications for both global and national security."

India joins broader cybersecurity initiative

The expansion comes weeks after reports suggested Indian government agencies and technology companies were evaluating participation in Project Glasswing.

According to earlier reports, organisations including TCS, Infosys and CERT-In were expected to gain access to the model to assess vulnerabilities across critical software systems.

While Anthropic has not disclosed the full list of participating organisations, the company said new entrants “must meet specific security and eligibility requirements” before receiving access.

Unlike general-purpose AI models that are primarily used for content generation, coding assistance or workflow automation, Claude Mythos is designed to support software assurance and cybersecurity testing activities.

The model is intended to help organisations identify weaknesses in software code, applications and digital infrastructure before they become security incidents.

Partner ecosystem takes shape

The expansion in India also comes as Anthropic increases its local enterprise and partner investments.

In February, the company formally opened its Bengaluru office, describing India as Claude's second-largest market globally.

At the time, Anthropic said its India run-rate revenue had doubled since its expansion announcement in late 2025, due to “growing enterprise adoption” across software development, application modernisation and operational AI deployments.

Companies including Air India, Razorpay and Cognizant have publicly discussed their use of Claude across engineering and business operations.

More recently, Anthropic appointed Siddiq Zaman as head of partnerships for India.

Zaman previously held partnership and alliance leadership roles across ServiceNow, AWS, Salesforce and HPE.

The appointment clearly signals Anthropic's increasing focus on building relationships with enterprise partners, consulting firms and system integrators as AI adoption moves beyond experimentation into production environments.

Potential opportunity for security partners

The arrival of Claude Mythos could create new opportunities for cybersecurity providers, managed security services firms and large system integrators.

With organisations adopting AI-assisted approaches to software development, there is growing pressure to improve software assurance, code quality and vulnerability management processes.

Cybersecurity testing, secure software development practices and vulnerability assessments have traditionally relied heavily on manual reviews and specialised tools.

AI models, including Claude Mythos, could become an additional layer within those workflows, helping organisations identify potential weaknesses earlier in the software development lifecycle.

For partners, the opportunity may not be limited to deploying AI models.

System integrators and cybersecurity providers could play a larger role in integrating these capabilities into software development pipelines, security operations processes and governance frameworks.

While the long-term impact remains to be seen, Anthropic's latest move suggests that its India strategy is expanding beyond enterprise AI adoption into cybersecurity and software resilience, areas where partners are likely to play a significant role.