AI company Anthropic is in talks with governments around the world, including India, about protecting critical infrastructure ahead of the public release of its Claude Mythos AI model, according to reporting published in April 2026.
The discussions involve India’s tech industry body Nasscom and the Indian government, both of which are seeking early access to the model. Their goal is to use that access to strengthen cybersecurity defenses before the model becomes publicly available.
The concern centers on the potential vulnerabilities that Claude Mythos could expose once released. The model is described as powerful enough that its public availability may create significant security risks for critical infrastructure.
Anthropic’s outreach to governments suggests the company is aware of those risks and is working proactively with national authorities to address them before a wider rollout. India’s involvement reflects the country’s interest in ensuring its digital and physical infrastructure is prepared for the implications the model could bring.
For India’s technology sector and government, early access to Claude Mythos could allow security teams to identify and address weaknesses before bad actors have the same opportunity. Nasscom’s participation indicates the private sector is also closely involved in these preparedness efforts.
The talks highlight a broader pattern of AI developers engaging with governments on safety and security questions as AI models grow more capable — though the specific terms or timeline of any access agreement between Anthropic and Indian stakeholders have not been disclosed in available reporting.
Source: Tech-Economic Times