Views of Ashwini Vaishnaw: “AI must be democratised”

India has embarked on a new phase in its technology journey, as focus shifts from building digital public infrastructure (DPI) at scale to making artificial intelligence (AI) available, affordable and useful for every citizen and sector. At the recently concluded India AI Impact Summit 2026, Ashwini Vaishnaw, Union Minister of Electronics and Information Technology, spoke about AI as a foundational technology, outlined India’s approach across the five layers of the AI stack, and emphasised outcomes-led adoption through real-world applications, sovereign models, common compute as a public good, clean energy and a techno-legal approach to contain harms. Edited excerpts from his remarks…

AI is foundational, focus is on delivery

AI is a foundational technology. It is already transforming how we work, learn and make decisions. The government’s vision is to democratise this technology, deploy it at scale and make it accessible to all. In India, we are working on all the five layers of the AI stack. We treat this as a complete ecosystem, and our focus is on getting benefits to the masses while building capabilities that allow us to control our own destiny.

The first layer is the application layer. We believe that return on investment (RoI) will come from this layer, and this is where the real benefit will flow to the masses. We are building real-world solutions in healthcare, agriculture, education, logistics, design and financial services, where the benefits can reach the masses. This is the fifth industrial revolution, and it will impact every sector of the economy and society.

The second layer is the model layer, and this is where sovereignty comes in. While very large models will continue to push the frontier, we believe that more than 90 per cent of use cases can be addressed with smaller, focused models. These specialised models will deliver greater value to our masses by addressing specific use cases at much lower cost. Our bouquet of sovereign models, many of which have been launched during the India AI Impact Summit 2026, is focused on multimodal and multilingual capabilities because we have a very diverse culture and a hugely diverse linguistic scenario in our country. Also, tech neutrality is important. Like sovereign AI, we should be able to control our own destiny.

The third layer is the compute layer. We treat compute as a public good. In a public-private partnership, we have created a common compute platform where we provide access to 38,000 graphics processing units (GPUs) at very affordable rates for start-ups, academia, researchers and students, and we will be adding another 20,000 GPUs to this common compute platform. Unlike in so many other countries, where the AI infrastructure is under the control of a handful of companies, we in India have been able to provide AI compute to a very large section of the population.

Just like DPI brings technology to vast sections of the population, we have created a similar structure as part of the IndiaAI Mission 1.0. We are also expecting another 20,000 GPUs over and above the 13,000 GPUs that we already have in the common compute layer. Going forward, we will be placing orders for another 20,000 GPUs, and they will be deployed within the next six months. It is a constant endeavour to provide high-quality resources to our start-ups, researchers and students. The government is focused on making technology accessible to people who want to use it.

The fourth layer is the infrastructure layer. With our talent pool and very trusted policies, and with the recent budget announcing a major policy shift towards getting the world’s data to India, to reside in India, to be processed here, and to deliver high-value services to the entire world, we expect large investments in data centres in the coming months.

The fifth layer is the energy layer. In the country today, more than 50 per cent of power generation capacity is from clean sources, such as renewables. The government has also recently reformed the nuclear energy sector, which is another good sector for providing the base load of clean energy. We also have a robust grid that has been practically rebuilt over the past one decade.

Investment is coming in all five layers of the AI stack. We have seen venture capital firms and other players committing funds for big solutions and big applications. We have also seen commitments for further research in cutting-edge models, because the next level of models will come from engineering innovation as well as mathematical innovation. We are seeing huge investment interest in the infrastructure layer and the energy layer as well.

The investment opportunity identified today is likely to exceed $200 billion over the next two years across the five layers of the AI stack. We can also now see strong commitment from venture capital firms to our deep tech start-ups.

Promoting responsible use and reducing risks

We are focusing on making AI accountable to society, ensuring that we harness the benefits of this new technology while containing the harms it may bring. We will try to create a consensus among global leaders about the good use, the proper use and the right use of AI, so that it can be used for the benefit of humanity while we contain the harms that may come from improper use. A good consensus is emerging among global leaders. Our interactions so far with many other countries show that everybody believes AI should be used for good, and all harmful impacts should be mitigated.

Adopting a balanced techno-legal regulatory framework

Containing harmful impacts must be done through a techno-legal approach. This cannot be achieved merely through regulation or by passing a law. At the AI Safety Institute, which is a virtual institute, we are working with many academic institutions to create technical solutions to prevent the harmful impacts of AI. We are conscious that while we harness the benefits of AI, we must also find collective solutions for mitigating the risks. By placing human safety and dignity at the heart of AI, we can move forward with conviction.

The Digital Personal Data Protection Act, 2023, is a well-balanced framework in which innovation and regulation are harmonised in a healthy manner. We would like to follow a similar structure in all our regulations. We will come out very heavy on harmful impact, and we will promote innovative capabilities and encourage the use of technologies for solving genuine problems of society.

Talent, reskilling and the next pipeline

The IT industry is one of India’s biggest strengths. Whenever any technology transition happens, it must be managed jointly by industry, academia and government. Right now, we are working with industry and academia on three fronts: reskilling and upskilling the existing talent in the workforce; creating a new talent pipeline; and making sure that the coming generations are ready for this new technology.

So far, more than 100 colleges are working with the IT industry to reskill talent. The Future Skills programme, which we started three years ago, is now deployed for AI-based reskilling. The Ministry of Education and the All India Council for Technical Education are working on changes in the course curriculum to update the talent pipeline and ensure that the youth receive new opportunities.

Strengthening semiconductors

Regarding chips, the India Semiconductor Mission (ISM) is cognisant of the advancements required for the AI world. As we go forward with ISM 2.0, design will be the first focus. We believe that some big developments will come out of India in the coming years because of the efforts that we are making.

Regarding memory, we will soon begin commercial production at a large facility that will produce memory chips. There are many more people who want to invest in memory manufacturing here in India. We are also taking up a research project as part of ISM 2.0 where we can develop intellectual property around the latest high bandwidth memory chips, and leading researchers are responding encouragingly to efforts aimed at developing the next level of high bandwidth memory in the country.

IndiaAI Mission 2.0 and the next phase

As we move forward, we will soon launch IndiaAI Mission 2.0. There will be a significant focus on research and development, innovation, AI diffusion, and further strengthening the common compute infrastructure developed under IndiaAI Mission 1.0. It is a constant endeavour to provide high-quality resources to start-ups, researchers and students.

We believe AI must be democratised. The AI divide is a technological divide, like any other social phenomenon. It is something that we must consciously try to bridge.