Economic Survey 2025-26 flags the need for India's own AI solutions

It said that India is among the top contributors to global AI research, has a large pool of technical talent, and has one of the most AI-aware workforces in the world

The Economic Survey 2025–26 has highlighted that India must develop its own artificial intelligence (AI) solutions instead of depending heavily on foreign companies. It said that AI is still at an early stage globally, giving India a chance to shape its future in a way that creates better jobs and long-term value for its workforce.

 

"AI is no longer a distant or speculative technology. It is increasingly being adopted, even if in an experimental capacity, in organisations around the world," the Survey said. It noted that by building indigenous AI capabilities, India can create more meaningful and dignified employment instead of remaining just a source of cheap labour.

India's approach towards AI

The Economic Survey said that the global AI world is moving in two different directions. In Western countries, big technology firms are following a top-down approach, focusing on powerful AI models, large private funding, expensive computing infrastructure and control of technology by a few companies.

 

In contrast, many other countries are using a bottom-up approach, where innovation is spread across sectors, supported by the government, and focused more on practical AI applications rather than building the biggest models. The Survey notes that India’s situation makes a bottom-up AI approach the right choice.

 

It said that India has strong advantages, including being among the top contributors to global AI research, a large pool of technical talent, and one of the most AI-aware workforces in the world. It also has access to rich and diverse domestic data from sectors such as health, agriculture, finance and public services, though this potential has not yet been fully used.

Lack of computing power a key challenge

The Survey pointed out that India needs a strong computing capacity to develop advanced AI models. However, rising demand for graphics processing units (GPUs), shortages of critical chips and storage, and high costs are creating major hurdles. Even if funding is available, data centre expansion may be delayed due to a lack of hardware supply.

 

The Survey noted that while financing challenges ease over time, access to GPUs becomes the biggest long-term bottleneck. Even when demand is strong and money is available, uncertainty in global GPU supply slows expansion.

Human capital for AI

The Survey highlighted that to build useful AI models for India, the country needs people with two key skills: a strong understanding of algorithms and good software engineering abilities. This knowledge mainly comes from hands-on experience and is not easily learned from books. The Economic Survey says India must attract experts who have built large AI models and encourage them to train others.

 

It also stresses that countries leading in innovation mix classroom learning with real industry exposure. For India, AI training should start early, at the school or university level, and be done with help from the private sector. Universities need the freedom to design flexible, industry-relevant courses, a move supported by the Viksit Bharat Shiksha Adhishthan Bill, 2025.