India is building a dedicated military AI layer to turn battlefield data into faster decisions while keeping sensitive defence information under domestic control. On October 23, Defence Secretary Rajesh Kumar Singh said the Ministry of Defence is working with the IndiaAI Mission to develop the capability for decision support, target acquisition and data fusion, including applications across the military’s sensor-to-shooter loop.
The idea is to give the Armed Forces greater control over the AI infrastructure processing operational data. Rajesh Kumar Singh said India cannot depend entirely on foreign companies for technologies where the underlying information is highly sensitive, although collaboration with trusted international partners will remain possible. He also linked the effort to a wider defence-policy shift from simply manufacturing equipment domestically to owning more of the designs, intellectual property and source code behind critical systems.
In practical terms, the sensor-to-shooter loop connects information collected by military sensors with analysis, target identification and operational response. AI can help combine data from multiple sources and support commanders in reaching decisions faster. The current reporting supports decision assistance and target acquisition; it does not establish that the proposed AI layer will independently select or engage targets without human authority.
India already has a national computing base that could support such development. The IndiaAI Mission, approved with an outlay of about ₹10,372 crore, had expanded its shared compute pool to more than 45,000 GPUs by June 2026. By August, 237 projects had consumed 93.18 lakh GPU hours through subsidised access, while the government had selected 20 indigenous foundation-model proposals from 506 applications. Rajesh Kumar Singh said compute for the military initiative is being secured through IndiaAI and that private companies already selected under the mission are expected to support development.
The military AI push also sits inside a broader effort to build more indigenous autonomous systems. In September, the Indian Air Force released its UAS Roadmap at Dronathon 2026, covering surveillance, autonomous operations, swarm technologies, electronic warfare and counter-drone systems. An October industry programme then identified indigenous propulsion, autonomous software, resilient navigation and standardised interfaces as priority technologies for reducing dependence on external suppliers.
India has also been working on how AI should be trusted inside defence systems, not only how it should be built. In 2024, the Ministry of Defence introduced the Evaluating Trustworthy Artificial Intelligence (ETAI) Framework and Guidelines for the Armed Forces, covering areas such as reliability, robustness and resilience against adversarial attacks. That framework provides an existing governance layer for integrating AI into sensitive defence operations.
Several important details remain undisclosed. The government has not publicly identified the AI models that will power the military layer, the private companies involved, the final architecture, a dedicated programme budget or a deployment timeline. “Military AI layer” should therefore be understood as the capability Rajesh Kumar Singh described rather than the confirmed name of an operational platform.
The larger shift is that India wants to own more than AI-enabled weapons or individual defence applications. It is trying to build domestic control over the computing, data and intelligence layer that can connect battlefield sensors with military decisions. If that architecture takes shape, the strategic advantage will come not simply from having more AI, but from controlling the technology through which sensitive battlefield information becomes actionable intelligence.
