The Indian government is currently in high-level discussions with Anthropic and the U.S. administration to secure early access to Claude Mythos, a “frontier” AI model that has sent shockwaves through the cybersecurity industry since its announcement on April 7, 2026.
According to reports from Tuesday, April 28, 2026, the Ministry of Electronics and Information Technology (MeitY) and the Finance Ministry are treating the model’s release as a matter of “national critical infrastructure security.”
1. What is Claude Mythos?
Anthropic describes Claude Mythos as a model whose cybersecurity and agentic coding capabilities are “substantially beyond” anything previously trained.
- The Cyber “Happy Accident”: Anthropic claims Mythos’s extreme hacking skills were an accidental byproduct of training it for advanced software engineering and multi-step reasoning.
- Zero-Day Specialist: In internal testing, Mythos discovered a 27-year-old bug in OpenBSD that had eluded decades of human and automated audits. It also successfully reversed-engineered exploits for modern browsers like Firefox 147.
- Capability Jump: While the previous top-tier model (Opus 4.6) found roughly 500 vulnerabilities in test repos, Mythos identified tens of thousands, representing a 20-fold generational leap.
2. Project Glasswing: The Exclusive “Defenders” Club
Because Mythos is deemed too dangerous for general release, Anthropic launched Project Glasswing—a gated research preview designed to let “defenders” patch their systems before the model (or its successors) potentially leaks to bad actors.
- The Current Roster: Anthropic granted early access to 40 major firms, including Microsoft, Google, Apple, Amazon (AWS), CrowdStrike, and JPMorgan Chase.
- The India Gap: Currently, no Indian companies are included in Project Glasswing. This has triggered a “red alert” within the Indian government and IT sector.
3. India’s Three-Pronged Negotiation
The Indian government is pushing for “equitable access” through multiple channels to ensure the country is not left vulnerable to an “AI Divide.”
- Bilateral Logistics: MeitY Secretary S. Krishnan confirmed on April 28 that the Centre is working out the “logistics” with U.S. authorities to include Indian entities in the program.
- NASSCOM Advocacy: The industry body has formally written to Anthropic, arguing that because Indian IT code runs much of the “global digital backbone,” excluding Indian engineers from Mythos access creates a massive global security risk, not just a national one.
- Banking Vigilance: Finance Minister Nirmala Sitharaman convened an emergency meeting with the RBI and top banks last week, describing the threat from Mythos as “unprecedented” and requiring immediate vigilance across financial channels.
4. Strategic Concerns for Indian IT
Beyond security, there is a significant economic fear that Mythos could disrupt the Indian IT services model.
- Agentic Deflation: Analysts warn that Mythos’s ability to handle autonomous coding tasks could accelerate “deflationary pressures” in the IT sector, as clients may stop paying for human-hour “effort” when AI agents can deliver “outcomes” instantly.
- Competitive Disadvantage: If global rivals like Accenture gain Mythos access while TCS, Infosys, and Wipro are locked out, Indian firms could struggle to compete in the new “agentic operating model” era.