India pushes for sovereign hosting of Anthropic’s AI models amid Claude Mythos cybersecurity concerns

Indian government officials are exploring ways to host Anthropic’s Claude AI models domestically to safeguard critical infrastructure, including banking networks, telecom systems, and power grids. The push follows cybersecurity concerns tied to the Claude Mythos system, which officials fear could expose key sectors to risks. Anthropic, the developer of Claude, has not guaranteed India-specific access or hosting arrangements, leaving the country’s ability to secure its infrastructure dependent on external assurances. The demand for sovereign hosting stems from broader efforts to protect national data and systems from potential cyber threats. Indian firms and government agencies rely on advanced AI models for operations, but concerns over foreign-hosted systems persist amid rising global cybersecurity threats. The move aligns with India’s broader strategy to reduce dependency on overseas cloud and AI services for sensitive applications. While no formal agreement has been reached, discussions are underway to ensure compliance with India’s data localization and cybersecurity regulations. The government is reportedly evaluating partnerships with local tech firms to facilitate secure, domestic AI deployments. Anthropic’s response remains unclear, but the situation highlights growing tensions between global AI accessibility and national security priorities. If unresolved, the issue could impact India’s adoption of cutting-edge AI tools in critical sectors.