The Indian government is likely to encourage the use of domestically developed artificial intelligence (AI) models across critical sectors, including defence and aerospace, as it looks to strengthen the country’s technological independence. The proposed push could involve asking key government departments at both central and state levels to prioritise Indian AI systems and avoid relying on advanced models and AI agents developed by foreign companies, according to a Business Standard report published on November 9, 2026, citing senior government officials.
The reported move is part of India’s broader effort to build its own AI ecosystem through the ₹10,372 crore IndiaAI Mission. The government is supporting the development of indigenous large language models (LLMs), sector-specific small language models (SLMs) and AI-powered applications. The objective is to ensure that Indian institutions have access to AI technologies that can operate in sensitive environments, address domestic requirements and reduce dependence on foreign technology providers.
Government May Prioritise Indian AI Models in Defence and Aerospace
According to the report, the government is considering encouraging critical departments to use AI systems developed by Indian companies, particularly in sectors where data security and technological control are important.
Defence and aerospace are among the sectors expected to receive attention. These areas often handle sensitive information, including operational data, technical documentation and strategic communications. Using AI systems developed and deployed domestically could provide government agencies with greater control over how such information is processed and stored.
The proposal could extend beyond defence and aerospace to other government-owned and government-operated critical sectors. Officials have reportedly discussed the possibility of asking departments at both central and state levels to avoid using frontier AI models and AI agents developed by non-Indian companies.
However, the reported proposal should not be interpreted as an officially announced blanket ban on foreign AI tools. The government has not publicly issued a final directive establishing such a restriction. The current development points to a possible policy preference for indigenous AI adoption in sensitive government operations.
Defence Departments Exploring Secure, Air-Gapped AI Systems
One of the major reasons behind the proposed shift is the need to operate AI systems in secure environments.
Some defence-related departments are already developing air-gapped large language model systems. These systems operate separately from the public internet and external networks, reducing the risk of unauthorised access to sensitive information.
Such models can also be designed to operate at the edge, allowing AI processing to take place closer to where data is generated or used rather than relying entirely on remote cloud infrastructure.
The government has reportedly asked some sectors to examine whether indigenous models developed under the IndiaAI Mission can be integrated into these existing systems.
This approach could be particularly useful for organisations that need AI assistance with document analysis, information retrieval, technical support and other internal workflows without sending sensitive information to external platforms.
IndiaAI Mission Supports Development of 20 Indigenous Models
The government’s push is closely connected to the IndiaAI Mission, which was approved by the Union Cabinet in March 2024 with a total financial outlay of approximately ₹10,372 crore over five years.
Under the mission, the government has identified and approved the development of 20 AI models. These include general-purpose foundation models, large language models and smaller models designed for specific industries and applications.
The projects are being developed by 12 selected startups and technology companies. Among the organisations involved are Sarvam AI and BharatGen, which have already released several models.
Other companies participating in the programme include Gnani, Gan AI, Fractal, Tech Mahindra and Shodh AI. Some of these companies are working on additional models that are expected to be released over the next 12 to 15 months, according to the officials cited in the report.
The government’s objective is to expand the capabilities of these models so they can serve demanding use cases across public administration, defence, industry and other important sectors.
Key details of India’s indigenous AI development programme
| Particular | Details |
|---|---|
| Government programme | IndiaAI Mission |
| Total outlay | ₹10,371.92 crore |
| Approved AI models | 20 |
| Selected development teams | 12 startups |
| Model categories | LLMs, foundation models and sector-specific SLMs |
| Key focus | Domestic AI capabilities and technological independence |
| Potential applications | Defence, aerospace, government and industry |
The IndiaAI Mission also supports access to computing infrastructure, datasets, startup financing, skills development and AI safety initiatives.
Indigenous AI Models Are Beginning to Reach the Market
India’s domestic AI development programme has already produced several models that could support the government’s ambitions.
On November 3, 2026, the Ministry of Electronics and Information Technology announced the release of EKA-1, developed by Soket Labs, and Dhīr 2B and Dhīr 5B, developed by ZenteiQ under the IndiaAI Mission.
EKA-1 is a foundation model with 26.3 billion parameters, trained on approximately 7.5 trillion tokens. It supports all 22 scheduled Indian languages as well as English and is designed for applications involving mathematics, coding, reasoning and agentic capabilities.
ZenteiQ’s Dhīr models are open-source multilingual assistants supporting English and 10 Indian languages. They are designed for secure, on-premise and offline-capable deployment.
The models could be relevant for government agencies and organisations that need AI capabilities without depending on continuous cloud connectivity.
Other companies, including Sarvam and BharatGen, are also developing domestic AI systems aimed at Indian languages and industry-specific requirements.
These developments indicate that India’s AI strategy is moving beyond research and infrastructure investment towards the practical deployment of locally developed models.
Why India Wants to Reduce Dependence on Foreign AI Companies
The growing interest in indigenous AI models is connected to several strategic and economic considerations.
Data security and sovereignty: Critical government departments handle sensitive information. Using domestically controlled systems could give institutions greater oversight of data storage, access and processing.
Reduced foreign technology dependence: Reliance on a small number of international AI providers can create strategic and operational risks. Developing domestic alternatives could give India more control over important technology infrastructure.
Support for Indian startups: Government adoption could create opportunities for Indian AI companies to secure institutional customers, generate revenue and improve their products through real-world deployment.
Local language capabilities: Indian AI developers can focus on regional languages, local administrative requirements and use cases that may receive less attention from global technology companies.
Customisation for critical industries: Smaller, specialised models can be developed for particular tasks and environments, potentially making them more suitable for specific government and industrial applications.
The Bigger Picture
India’s proposed preference for indigenous AI models reflects a broader shift towards technological self-reliance. Just as the country has invested in domestic digital infrastructure, semiconductor manufacturing and telecommunications, it is now trying to develop greater control over the AI systems that may become central to government operations and strategic industries.
However, building domestic AI models is only one part of the challenge. Their reliability, accuracy, security, computing requirements and ability to perform consistently in real-world environments will determine whether they can replace or complement established international systems. The reported claim that some domestic models have reached up to 80 per cent of the capabilities of leading frontier models also highlights the need to evaluate performance across specific tasks rather than rely on a single broad comparison.
The policy direction could provide Indian AI startups with an important opportunity, particularly if government departments become early adopters of their products. At the same time, procurement decisions will need to balance domestic technology preferences with performance, safety, cost and operational requirements.
Looking Ahead
The next stage will depend on how the government translates its reported preference for indigenous AI into practical procurement guidelines and deployment policies. Critical departments may evaluate models developed under the IndiaAI Mission before deciding which systems can be used in sensitive environments. Any formal restrictions on foreign AI systems would require clear government instructions, which have not yet been announced in the reported development.
For Indian AI companies, greater adoption by government institutions could become a significant source of demand and an opportunity to improve their models. The success of the initiative will ultimately depend on whether homegrown systems can deliver the security, reliability and performance required by defence, aerospace and other critical sectors while building a sustainable domestic AI industry.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
