MeitY: Halt OpenAI, Anthropic AI for Cybersecurity

A department under India's Ministry of Electronics and Information Technology has advised other ministries to temporarily halt the deployment of OpenAI and Anthropic models for cybersecurity. This guidance comes after both companies approached several ministries with proposals for their AI models. An official memorandum has been circulated, urging ministries not to use these models prematurely. The Ministry of Finance, for example, had specifically inquired about deploying OpenAI's GPT-5.5 for cybersecurity. The department's response was a firm "no" last week. However, this is not a permanent restriction; the concern is about the timing of deployment, not a ban on the technology itself. These AI models are designed for tasks like automated vulnerability discovery and incident triage. This decision is part of a broader government push for a sovereign AI agenda.