India's AI strategy should be application-led over frontier model development, says Economic Survey 2026

artificial intelligence

• India's AI strategy should prioritise application-led innovation over frontier model building.
• Survey urges bottom-up approach with open-source, sector-specific AI models
• Limited resources make foundational model development challenging for India

Did our AI summary help?

In the artificial intelligence (AI) era, India's advantage lies in application-led innovation rather than replicating frontier scale model development, the Economic Survey released on January 29 said.

While the efforts provide valuable process knowledge, the country's true strengths are its domestic data, human capital, and the ability of public institutions to coordinate large-scale efforts, it noted.

These remarks come amid an ongoing debate in India's tech and startup ecosystem on whether the country should focus on foundational AI or the application layer.

Bottoms-up approach in AI

"A bottom-up strategy anchored in open and interoperable systems, sector-specific models, and shared physical and digital infrastructure offers a more credible pathway to value creation than a narrow pursuit of scale for its own sake," the survey noted.

This proposed approach prioritises application-specific, small models that are tailored to defined uses and sectoral needs. The survey notes these models are significantly more computationally efficient, easier to fine-tune, and capable of running on locally available hardware, such as smartphones or personal computers, making them better suited to India’s existing infrastructure base.

More importantly, they allow innovation to emerge from a broader set of stakeholders, including startups, research institutions, public agencies, and domain-specific firms and allows broader development and diffusion of AI solutions without any concerns of resource constraints or high entry barriers.

Encourage open-source platforms

The Survey notes that the strategy must also encourage innovation on open-source and open-weight platforms which will enable the country to achieve more with less resources.

"India is currently one of the world’s largest and fastest-growing communities of open-source developers. They are active contributors to global codebases, often working in collaborative environments. Unifying these efforts under the scope of the IndiaAI Mission is essential to steer the potential of our talent pool towards shared domestic innovation," the survey noted.

Providing India's talent a robust platform and policy guidance can help reduce the country’s dependence on foreign proprietary systems, lower entry barriers for domestic developers, and create an environment that incentivises experimentation across sectors at a relatively low cost, the survey stated.

AI as public good

It further noted that the approach to coordinate a bottoms-up strategy can be spearheaded under an ‘AI-OS’ initiative, similar to UPI and Aadhaar, thereby turning AI into a public good.

"This will enable the sovereign to collaborate with state and local institutions to expand the availability of structured, anonymised, and machine-readable datasets in priority sectors; pool existing data centre capacity to create shared cloud compute infrastructure; and establish common platforms where open-source AI efforts can be coordinated and audited" the document read.

The survey noted that sector leaders in the West have pursued a top-down strategy centred on frontier models, massive private capital, significant expenditures on computing infrastructure, and the concentration of intellectual property within a small number of hyperscale firms.

India’s limited access to cutting-edge compute infrastructure, scarce financial resources for large-scale model training, and relatively muted private participation in foundational AI research compared to global leaders, however, makes pursuit of foundational models as the centrepiece of an AI strategy challenging. This makes a bottom-up approach to AI development align more closely with India's realities, the survey noted.

Debate on applications vs frontier model development 

Many tech stalwarts like Infosys chairman Nandan Nilekani have previously said that India should be the AI use-case capital of the world, rather than focusing on the hypercompetitive race of building Large Language Models (LLMs). Google DeepMind chief Demis Hassabis and Meta's global public policy chief Joel Kaplan have also backed a focus on application layer.

Entrepreneurs like Sarvam AI's Vivek Raghavan however have argued that without sovereign AI models, the country will become a "digital colony".

Global AI pioneers like Yoshua Bengio have previously highlighted the geopolitical implications of AI. "Countries that invest in building their AI systems will have a geopolitical advantage," Bengio told Moneycontrol in January 2025, adding that there’s a need for governments to fund and incentivise foundational model development.

Over the past year, the Indian government has selected around 12 players including Sarvam, Tech Mahindra, Gnani.ai, and Gan AI to build foundational large language models. Some of these startups are expected to showcase their models in the India AI Impact Summit 2026 next month.