Sarvam AI signs framework agreement with Integrated Defence Staff to build AI for all three services
Sarvam AI has signed an MoU with Headquarters Integrated Defence Staff creating an umbrella framework for AI projects across the Army, Navy, Air Force and tri-service organisations. The work will be delivered by Chanakya, Sarvam's firewalled national-security division, using its 105-billion-parameter foundation model.
Sarvam AI has entered a memorandum of understanding with Headquarters Integrated Defence Staff, the tri-service organisation under the Ministry of Defence, to design, develop and deploy AI systems for India's armed forces. The agreement establishes an umbrella framework rather than a single project, covering the Army, Navy, Air Force and joint and integrated organisations under HQ IDS.
According to Sarvam, the scope spans operational, training and decision-support applications, with follow-on technical support included. The work will be run through Chanakya, a firewalled division the company has set up specifically for national security and defence, and will draw on the full Sarvam stack including its Sarvam 105B foundation model.
The announcement generated some confusion. Initial reports described a deal signed directly with the Ministry of Defence; Sarvam issued a clarification the following day that the counterparty is HQ IDS, while noting the ministry is its parent body. Neither HQ IDS nor the ministry has published its own statement, and no contract value or duration has been disclosed.
The timing aligns with Sarvam's reported fundraising talks at a valuation of up to $4 billion and with a broader push by the government to procure AI from domestic suppliers for sensitive workloads. Defence is one of the few sectors where a sovereign model is a hard requirement rather than a preference.
Why it matters
This is the first publicly announced framework agreement between an Indian foundation-model company and the armed forces. It establishes a template, a ring-fenced division working under an umbrella MoU, that other suppliers will likely be asked to follow, and it confirms that the defence establishment intends to adopt large language models for decision support rather than only for back-office tasks.
Indicops view
Defence and internal-security deployments impose constraints that commercial projects rarely face: air-gapped environments, classified data, and accountability for every output. The lesson for civilian departments is the same in milder form. We build every government deployment on the assumption that it will one day be audited, which means logged prompts, versioned models, documented evaluation and a human owner for each decision the system influences.
More in Government AI
Back to the newsroomDelhi wants critical departments to lean on India-built AI models, officials say
Senior officials say the central government wants departments in sectors such as defence and aerospace to use more AI built in India under the IndiaAI Mission, and may steer some away from non-Indian frontier models for critical applications. The details below are drawn from the primary source, with analysis labelled separately.
Pune's Froncort.AI wins $2.3 million from Google.org to take AI-guided TB screening to eight districts
Pune healthcare AI company Froncort.AI has received $2.3 million from Google.org to scale mAI-NETRA, an offline, local-language tool that directs frontline health workers to high-risk areas. A Pune pilot screened far fewer people per confirmed TB case. The details below are drawn from the primary source, with analysis labelled separately.
Andhra Pradesh clears Robotics and Physical AI Policy 2026-36 and land for a Reliance AI data centre
Andhra Pradesh's cabinet has cleared a ten-year Robotics and Physical AI Policy, allotted about 25 acres at Bhogapuram to Reliance for a gigascale AI data centre and approved a Rs 570 crore Epack unit. Incentive details are yet to be published.