MeitY Secretary S. Major opportunities for AI in jobs and governance, says Krishnan

We are less than a week away from the India AI Impact Summit, which will be attended by representatives from dozens of countries. Can you tell us where we are on AI from the Indian perspective?

We have taken an approach where we will try to provide three aspects of the infrastructure that AI needs: compute, datasets, and models. With government support, access to these has become a little easier. Then the focus is on seeing what we can do with the applications and solutions that people are able to develop using these resources.

Ultimately, there are two things that are important. One, companies’ revenues will depend on how they deploy AI. The deployment is important and so is the impact. In the Indian context, there are many areas where you can use AI to increase productivity, efficiency and effectiveness. Our start-ups can do well and these are things that we can also offer to the rest of the world as products.

Given the kind of resources we have, we have to do it a little frugally, which is why the model we have adopted appeals to many countries in poorer parts of the world. On many of the AI-related indicators that Stanford University and other institutions measure, we are performing relatively well. On the Vibrancy Index, we are ranked third, on skills penetration and use of AI for enterprise solutions, we are ranked second overall.

So, if you look at this kind of penetration and this kind of skill, clearly we see some advantage there that we need to pursue. NITI Aayog has done a study which shows that yes, undoubtedly we will lose, or go away, some jobs in the routine coding programming side of IT/ITeS (Information Technology-Enabled Services), but we can also create more jobs and what if.

What I really see as a big other opportunity is that there are a number of areas, including governance, where we would all like to see a substantial increase in quality and potentially that’s something AI can offer. At the same time, we are aware of the risks, dangers and potential harms, which is why I think we are ready to regulate when regulation is needed.

What does regulation look like in practice?

If you have seen the report chaired by the Principal Scientific Advisor on AI governance guidelines, it also says try to use existing laws as much as possible. For example, if you look at what we can do with the existing Information Technology Act, this is one aspect of it. The second part is what we need to do in the copyright area. So, he is being dealt with in a special way. The second part is how other data, including personal data, is used. So, the Digital Personal Data Protection Act, 2023 fits in there.

Some of this regulation is already in place. Some of this requires change, strictness, and that is what we keep trying to do, including the new set of rules we have imposed (amendment in IT rules, requiring labeling of artificially generated content in 2021).

Those rules introduce labeling for AI-generated content and reduce the takedown timeline for all content from 24-36 hours to two to three hours.

Labeling is in the context of the right to know. We all have the right to know whether what we are seeing is artificially generated or not. This is a very small requirement and technically quite easy to solve. There were some issues which I think they (stakeholders) raised with us during the consultation (from October 2025, when the draft of these rules were published) and we have addressed them. For example, we have discounted smartphone camera auto-enhancement. Similarly for special effects in films.

The change in time frame is basically based on our understanding that there are two factors involved. When these timelines were initially implemented, they were much longer because the nature and type of intermediaries we were dealing with in those days were different and they had more time to respond.

The potential virulence of a lot of these things goes very quickly. All damage occurs within 24 or 36 hours. In practice, our own experience has been that whenever such a removal is required, most companies do not require more than an hour or two to comply.

On electronics manufacturing, how prepared are we in the age of weaponized supply chains?

Some of the story is in the past, some in the future. We also produced electronics until the late 1990s. Much of it moved out following the 1997 Information Technology-I Agreement (which allowed IT hardware to be imported at minimal duty). I’m not saying for a moment that he was necessarily bad.

I think the IT revolution probably wouldn’t have happened if you didn’t have access to computers, laptops and various other devices on the scale that we did, thanks to opening up. Now, you have reached a level where I think it is important to have that capability domestically as well. We believe that it is a global value chain, so it is not that every part of it will be in India, but to ensure that the value chain is deep, you have to have a large part of it.

So, we kind of start from the finished product end (like smartphones) because that gives you scale and employment. Value addition in the country is only 18-20% as companies import most of the components. However, this is changing with schemes such as the Electronics Component Manufacturing Scheme, which encourages technology transfer, as China has learned from the Apple ecosystem. The scheme is expected to significantly increase value addition by 35-40%, which is comparable to China’s 40-50%. Semiconductors are more strategic and lower priced; It’s about what we are capable of doing. There is a Tamil saying, ‘Veralukketha Veekam (Don’t bite off more than you can chew)’. So, the question is, ‘How do you chew something you can’t bite and control?’ The Bharat Semiconductor Mission is designed based on what we can actually chew. We are not in the leading position. But we are in areas where there is still a lot of consumption and will remain so for the foreseeable future. The support should be extended for at least a decade, which is why India Semiconductor Mission 2.0 was also announced in the Union Budget. Therefore, we must move forward and move forward systematically, then move towards more leading edges.

There are reports of the deadline for compliance with the Digital Personal Data Protection Act, 2023 being reduced from 18 to 12 months. Why?

We haven’t shortened it. We have started consultations with the industry. The feedback we received was that the 18-month period is too long and there are many elements that companies are already prepared to comply with. So, can we actually talk to the industry and see if we can get that timeframe down? So, this is the context in which we are talking to the industry.

Varghese K. George: An international commentator compared the situation on AI now to our awareness of COVID-19 in February 2020. So, everyone was seeing some distant virus in China and then three weeks later everything in the world turned upside down. So, the comment is that that moment has already arrived in terms of AI. So, what is our understanding of where global AI research stands?

While much has been said about agentic AI taking over, our view is that its practical utility remains uncertain. We believe that focusing on smaller, specific AI tools – such as domain-specific, vision, quantitative models, and small language models – provides more immediate, practical relevance and greater benefits to society and humanity. Agentic vision may emerge, but it is still a long way off.

Jacob Koshy: How are IT companies discussing the AI ​​wave? Their business model is built on labor intermediation which is now being threatened by this technology.

We have spoken to many people from the IT industry. They say that many coding and programming jobs are difficult to maintain because they can be done by AI bots. But when you have to build an application, or create a solution, you need to have better domain expertise, like in agriculture or manufacturing. Deploying applications takes human resources. You have to understand what data sets you need to bring in, how you prepare them to suit a particular situation, how you adjust the way the orchestration tiers work, and many deployment-related tasks that need to be done. Their understanding is that they will still have many job opportunities. But this will require many of their current employees to be retrained and understand it differently. We have this program called Future Skills Prime, which is primarily designed to re-skill and re-train people. In colleges there has been an emphasis on teaching it as a horizontal technique; We have to teach it in every curriculum.

Suhasini Haider: Two questions – are we looking to create an international body for AI ethics and safety? And on MeitY’s cyber law division: Its purpose is to prevent unlawful speech and yet time and again we see people in government putting up AI videos inciting violence. Where do you think MeitY’s responsibility actually lies?

This is the first time a country from the Global South is hosting an AI summit. So, in a sense, yes, India could potentially be a natural leader in some aspects of AI, not necessarily in AI governance or regulation – that’s a part of it – but also in terms of offering more affordable technologies and more affordable deployment. Hopefully something will come out in the final announcement. Now, whether there will be another international organization like the Solar Alliance, I really don’t know. We can’t do it as a regular body – we’re also part of the UN Global Digital Compact and so on. So, we will work with the international community to see how this progresses. The number of cases where governments block information online is actually a fraction – it’s less than 0.1% of the total number of cases that social media entities actually remove as part of their community guidelines and so on. So, it’s very small, but when things come out through this channel we have to take action and we take action on the content that is brought to us.

Mr. Sampath: AI is a power-intensive sector with water and electricity needs. How are we looking at this with our climate commitments?

India has one of the largest grids in the world with high levels of renewable energy and load capacity. One problem with renewable energy is that there is no consumption at the time it is generated because the load is insufficient and most of it is returned. Therefore, there is an understanding that there may be excess electricity that can be used for this purpose. There are both air-cooled servers and water-cooled servers, and there are ways to make it affordable as well.

But we are absolutely clear that there is no relaxation given to any environmental norms or any other norms for the data centre. The only set of norms that have been relaxed are building construction norms; Data centers do not require much parking etc. To that limited extent, it is an exemption.

But in case of water and electricity consumption, they have to meet all the relevant norms depending on the availability, which is required to be done. Many of these decisions are ultimately taken at the state government level. There has not been much open promotion of data centers in all locations.