India’s push for sovereign artificial intelligence has moved firmly into the defence domain. Headquarters Integrated Defence Staff (HQ IDS) and Bengaluru-based Sarvam AI have signed a Memorandum of Understanding (MoU) to jointly develop secure, India-controlled AI capabilities for the Indian Army, Indian Navy and Indian Air Force.

Under the partnership, the two sides will collaborate on AI-based applications, research, knowledge management, training and decision-support systems tailored to the needs of the armed forces. The emphasis, according to details shared around the signing, is on capabilities that are secure, remain under national control and can be deployed at scale.

Data sovereignty at the core

A central pillar of the collaboration is data sovereignty — ensuring that sensitive defence data stays under India’s control — along with a commitment to responsible AI. For military users, where operational information cannot be routed through foreign-hosted models or platforms, an indigenous AI stack is not just a preference but a strategic requirement.

The MoU signals a broader shift in how the tri-services intend to adopt AI: not as off-the-shelf tools, but as systems designed, developed and governed with India’s data security and defence requirements in mind.

Building on DRONA 2.0

The agreement builds on an existing relationship. Earlier this year, the Defence Services Staff College (DSSC), Wellington, rolled out DRONA 2.0, an indigenous multimodal AI system for defence education, research and strategic studies that integrates Sarvam’s 105-billion-parameter model with open-source large language models, running on DSSC’s own GPU infrastructure. That platform was audited and certified by HQ IDS and the Defence Cyber Agency before deployment.

The new HQ IDS MoU widens that scope from a single institution to the joint-services level, opening the door to AI use cases across training, institutional knowledge management and command decision support.

Why it matters

The partnership comes as the Ministry of Defence and the IndiaAI Mission work towards a secure AI layer for the military, and as policymakers sharpen their focus on homegrown foundation models. Sarvam, selected under the IndiaAI Mission to build an indigenous foundational model, is increasingly positioning its Indian-built models for high-trust, public-sector deployments.

For India’s AI ecosystem, defence is among the most demanding test beds — requiring air-gapped or tightly controlled deployments, rigorous security audits and multilingual capability. Success here could set the template for sovereign AI adoption across other critical sectors, from public services to critical infrastructure.

As conversations on sovereign AI, responsible deployment and national AI capability take centre stage, partnerships such as this one underline how quickly Indian-built AI is moving from labs and pilots to mission-critical institutions.

Be a part of Elets Collaborative Initiatives. Join Us for Upcoming Events and explore business opportunities. Like us on Facebook , connect with us on LinkedIn and follow us on Twitter.

"Exciting news! Elets technomedia is now on WhatsApp Channels Subscribe today by clicking the link and stay updated with the latest insights!" Click here!