India negotiates access to Anthropic's Mythos AI model for cybersecurity

The Indian government is actively pursuing access to Anthropic's advanced AI model, Mythos, to strengthen its cybersecurity capabilities. IT Secretary S. Krishnan said talks are underway with US authorities and Anthropic, while the Indian Computer Emergency Response Team (CERT‑In) is already using alternative AI models in a dedicated sandbox to detect and fix software vulnerabilities. The effort is described as a “dry run” to prepare processes for when full access to Mythos becomes available.

Separately, the Ministry of Electronics and Information Technology (MeitY) has issued a memorandum to central ministries urging them to pause the deployment of OpenAI and Anthropic models for cybersecurity and related functions. The memo highlights dual‑use risks and the need for a clear regulatory framework, noting that ministries, including finance, have sought clarification on the use of agentic AI such as GPT‑5. The department says the pause is temporary pending further assessment and possible legislation.