Before Mythos goes public, Indian IT also wants access

Artificial intelligence (AI) company Anthropic is engaged in discussions between the US and allied democracies, including India, on securing critical infrastructure before the public release of its Claude Mythos model, said people with knowledge of the matter. Mythos, said to be the most powerful AI model developed till date, is expected to expose deep-seated vulnerabilities in the infrastructure of companies globally. Anthropic has held back a wider launch due to this fear while giving early access to a select group of companies to fix their systems. No Indian companies figure on this list.

Nasscom, representing India’s technology companies, has written to Anthropic, asking that they be included in Project Glasswing and be given access to Mythos to build cyber resilience since their code is used by companies across the globe. The Ministry of Electronics and Information Technology (MeitY) is also reportedly in discussions with Anthropic executives in the US on giving early access to Indian companies.

The conversations involving the governments and Anthropic focus on three areas--timing, scope and the structure of international access. The aim is to identify and plug vulnerabilities in global software and any cross-border risk that may emerge before the model is generally available.


Anthropic did not respond to ET’s queries till press time.

Nasscom did not confirm the development, but stated that India's technology services industry is deeply embedded in the global digital backbone, managing critical infrastructure, sensitive data and core operations across sectors and geographies.

“As AI systems evolve to autonomously identify and chain vulnerabilities across platforms, the potential for cascading, cross-border risk becomes significantly higher,” it said in a statement. “Given this reality, it is imperative that Indian technology firms are included in the global industry consortium by Anthropic.”

On Thursday, finance minister Nirmala Sitharaman convened a meeting with heads of various banks, the Reserve Bank of India and MeitY to understand the potential cybersecurity risks associated with Mythos.

The finance ministry said on X that the threat from Mythos was “unprecedented and requires a very high degree of vigilance, preparedness and better coordination across financial institutions and banks.”

Last month, Anthropic announced Project Glasswing, releasing Mythos to a group of 40 American companies including Microsoft, Amazon Web Services (AWS), Google, Microsoft, Apple, CrowdStrike, Palo Alto Networks, NVIDIA, Broadcom, Cisco, JPMorganChase and the Linux Foundation to help them detect and patch zero-day vulnerabilities.

As part of its Responsible Scaling Policy, Anthropic has given a headstart to US-based “defenders” and is also in discussion with top IT companies to be included in Project Glasswing.

If US competitors like Accenture and Capgemini gain access and Indian companies such as Tata Consultancy Services, Wipro and Infosys don’t, the latter would be at a competitive disadvantage, as the model’s use goes beyond cyber exploits, experts said.

Early testing has shown the model beats software engineering benchmarks by a wide margin, potentially making it the most powerful coding assistant ever and further threatening the IT business, which has already reported AI-led deflation in FY26. Companies have asked Anthropic for controlled access to prepare for what’s coming.

Infosys chief executive Salil Parkekh said however that models like Mythos open up opportunities.

“My sense is, if we build a good capability in that, we could help our clients to say, look, let's make sure that your vulnerabilities are better and quickly protected,” he said on an earnings call this week.

Forrester principal analyst Biswajeet Mahapatra said, “Restricted access to a frontier model like Mythos constitutes a severe competitive disadvantage by creating an AI Divide that stalls an organisation's transition to an agentic operating model.”

Mythos could be a primary catalyst for the deflationary pressures currently observed in FY26.

“This shift forces a fundamental breakdown of the traditional time and material revenue model, as clients increasingly reject paying for human-hour ‘effort’ when autonomous agents can deliver superior ‘outcomes’ instantly,” Mahapatra said.